refactor(skill): transition from artifact set to bundle structure

- Replaced SkillArtifactSet with SkillBundle across various components, enhancing the organization of skill dependencies and references.
- Updated SkillManager methods to load and save bundles instead of artifacts, improving clarity in asset management.
- Refactored SkillCompiler to compile skills into bundles, streamlining the dependency resolution process.
- Adjusted DifyCli and SandboxBashSession to utilize ToolDependencies, ensuring consistent handling of tool references.
- Introduced AssetReferences for better management of file dependencies within skill bundles.

.agents/skills/skill-creator/scripts/quick_validate.py

@@ -4,7 +4,6 @@ Quick validation script for skills - minimal version
 """
 
 import sys
-import os
 import re
 import yaml
 from pathlib import Path

.agents/skills/vercel-react-best-practices/AGENTS.md

@@ -33,11 +33,13 @@ Comprehensive performance optimization guide for React and Next.js applications,
    - 2.4 [Dynamic Imports for Heavy Components](#24-dynamic-imports-for-heavy-components)
    - 2.5 [Preload Based on User Intent](#25-preload-based-on-user-intent)
 3. [Server-Side Performance](#3-server-side-performance) — **HIGH**
-   - 3.1 [Cross-Request LRU Caching](#31-cross-request-lru-caching)
-   - 3.2 [Minimize Serialization at RSC Boundaries](#32-minimize-serialization-at-rsc-boundaries)
-   - 3.3 [Parallel Data Fetching with Component Composition](#33-parallel-data-fetching-with-component-composition)
-   - 3.4 [Per-Request Deduplication with React.cache()](#34-per-request-deduplication-with-reactcache)
-   - 3.5 [Use after() for Non-Blocking Operations](#35-use-after-for-non-blocking-operations)
+   - 3.1 [Authenticate Server Actions Like API Routes](#31-authenticate-server-actions-like-api-routes)
+   - 3.2 [Avoid Duplicate Serialization in RSC Props](#32-avoid-duplicate-serialization-in-rsc-props)
+   - 3.3 [Cross-Request LRU Caching](#33-cross-request-lru-caching)
+   - 3.4 [Minimize Serialization at RSC Boundaries](#34-minimize-serialization-at-rsc-boundaries)
+   - 3.5 [Parallel Data Fetching with Component Composition](#35-parallel-data-fetching-with-component-composition)
+   - 3.6 [Per-Request Deduplication with React.cache()](#36-per-request-deduplication-with-reactcache)
+   - 3.7 [Use after() for Non-Blocking Operations](#37-use-after-for-non-blocking-operations)
 4. [Client-Side Data Fetching](#4-client-side-data-fetching) — **MEDIUM-HIGH**
    - 4.1 [Deduplicate Global Event Listeners](#41-deduplicate-global-event-listeners)
    - 4.2 [Use Passive Event Listeners for Scrolling Performance](#42-use-passive-event-listeners-for-scrolling-performance)
@@ -568,7 +570,158 @@ The `typeof window !== 'undefined'` check prevents bundling preloaded modules fo
 
 Optimizing server-side rendering and data fetching eliminates server-side waterfalls and reduces response times.
 
-### 3.1 Cross-Request LRU Caching
+### 3.1 Authenticate Server Actions Like API Routes
+
+**Impact: CRITICAL (prevents unauthorized access to server mutations)**
+
+Server Actions (functions with `"use server"`) are exposed as public endpoints, just like API routes. Always verify authentication and authorization **inside** each Server Action—do not rely solely on middleware, layout guards, or page-level checks, as Server Actions can be invoked directly.
+
+Next.js documentation explicitly states: "Treat Server Actions with the same security considerations as public-facing API endpoints, and verify if the user is allowed to perform a mutation."
+
+**Incorrect: no authentication check**
+
+```typescript
+'use server'
+
+export async function deleteUser(userId: string) {
+  // Anyone can call this! No auth check
+  await db.user.delete({ where: { id: userId } })
+  return { success: true }
+}
+```
+
+**Correct: authentication inside the action**
+
+```typescript
+'use server'
+
+import { verifySession } from '@/lib/auth'
+import { unauthorized } from '@/lib/errors'
+
+export async function deleteUser(userId: string) {
+  // Always check auth inside the action
+  const session = await verifySession()
+  
+  if (!session) {
+    throw unauthorized('Must be logged in')
+  }
+  
+  // Check authorization too
+  if (session.user.role !== 'admin' && session.user.id !== userId) {
+    throw unauthorized('Cannot delete other users')
+  }
+  
+  await db.user.delete({ where: { id: userId } })
+  return { success: true }
+}
+```
+
+**With input validation:**
+
+```typescript
+'use server'
+
+import { verifySession } from '@/lib/auth'
+import { z } from 'zod'
+
+const updateProfileSchema = z.object({
+  userId: z.string().uuid(),
+  name: z.string().min(1).max(100),
+  email: z.string().email()
+})
+
+export async function updateProfile(data: unknown) {
+  // Validate input first
+  const validated = updateProfileSchema.parse(data)
+  
+  // Then authenticate
+  const session = await verifySession()
+  if (!session) {
+    throw new Error('Unauthorized')
+  }
+  
+  // Then authorize
+  if (session.user.id !== validated.userId) {
+    throw new Error('Can only update own profile')
+  }
+  
+  // Finally perform the mutation
+  await db.user.update({
+    where: { id: validated.userId },
+    data: {
+      name: validated.name,
+      email: validated.email
+    }
+  })
+  
+  return { success: true }
+}
+```
+
+Reference: [https://nextjs.org/docs/app/guides/authentication](https://nextjs.org/docs/app/guides/authentication)
+
+### 3.2 Avoid Duplicate Serialization in RSC Props
+
+**Impact: LOW (reduces network payload by avoiding duplicate serialization)**
+
+RSC→client serialization deduplicates by object reference, not value. Same reference = serialized once; new reference = serialized again. Do transformations (`.toSorted()`, `.filter()`, `.map()`) in client, not server.
+
+**Incorrect: duplicates array**
+
+```tsx
+// RSC: sends 6 strings (2 arrays × 3 items)
+<ClientList usernames={usernames} usernamesOrdered={usernames.toSorted()} />
+```
+
+**Correct: sends 3 strings**
+
+```tsx
+// RSC: send once
+<ClientList usernames={usernames} />
+
+// Client: transform there
+'use client'
+const sorted = useMemo(() => [...usernames].sort(), [usernames])
+```
+
+**Nested deduplication behavior:**
+
+```tsx
+// string[] - duplicates everything
+usernames={['a','b']} sorted={usernames.toSorted()} // sends 4 strings
+
+// object[] - duplicates array structure only
+users={[{id:1},{id:2}]} sorted={users.toSorted()} // sends 2 arrays + 2 unique objects (not 4)
+```
+
+Deduplication works recursively. Impact varies by data type:
+
+- `string[]`, `number[]`, `boolean[]`: **HIGH impact** - array + all primitives fully duplicated
+
+- `object[]`: **LOW impact** - array duplicated, but nested objects deduplicated by reference
+
+**Operations breaking deduplication: create new references**
+
+- Arrays: `.toSorted()`, `.filter()`, `.map()`, `.slice()`, `[...arr]`
+
+- Objects: `{...obj}`, `Object.assign()`, `structuredClone()`, `JSON.parse(JSON.stringify())`
+
+**More examples:**
+
+```tsx
+// ❌ Bad
+<C users={users} active={users.filter(u => u.active)} />
+<C product={product} productName={product.name} />
+
+// ✅ Good
+<C users={users} />
+<C product={product} />
+// Do filtering/destructuring in client
+```
+
+**Exception:** Pass derived data when transformation is expensive or client doesn't need original.
+
+### 3.3 Cross-Request LRU Caching
 
 **Impact: HIGH (caches across requests)**
 
@@ -605,7 +758,7 @@ Use when sequential user actions hit multiple endpoints needing the same data wi
 
 Reference: [https://github.com/isaacs/node-lru-cache](https://github.com/isaacs/node-lru-cache)
 
-### 3.2 Minimize Serialization at RSC Boundaries
+### 3.4 Minimize Serialization at RSC Boundaries
 
 **Impact: HIGH (reduces data transfer size)**
 
@@ -639,7 +792,7 @@ function Profile({ name }: { name: string }) {
 }
 ```
 
-### 3.3 Parallel Data Fetching with Component Composition
+### 3.5 Parallel Data Fetching with Component Composition
 
 **Impact: CRITICAL (eliminates server-side waterfalls)**
 
@@ -718,7 +871,7 @@ export default function Page() {
 }
 ```
 
-### 3.4 Per-Request Deduplication with React.cache()
+### 3.6 Per-Request Deduplication with React.cache()
 
 **Impact: MEDIUM (deduplicates within request)**
 
@@ -784,7 +937,7 @@ Use `React.cache()` to deduplicate these operations across your component tree.
 
 Reference: [https://react.dev/reference/react/cache](https://react.dev/reference/react/cache)
 
-### 3.5 Use after() for Non-Blocking Operations
+### 3.7 Use after() for Non-Blocking Operations
 
 **Impact: MEDIUM (faster response times)**
 
@@ -1715,79 +1868,32 @@ Micro-optimizations for hot paths can add up to meaningful improvements.
 
 **Impact: MEDIUM (reduces reflows/repaints)**
 
-Avoid changing styles one property at a time. Group multiple CSS changes together via classes or `cssText` to minimize browser reflows.
+Avoid interleaving style writes with layout reads. When you read a layout property (like `offsetWidth`, `getBoundingClientRect()`, or `getComputedStyle()`) between style changes, the browser is forced to trigger a synchronous reflow.
 
-**Incorrect: multiple reflows**
+**Incorrect: interleaved reads and writes force reflows**
 
 ```typescript
 function updateElementStyles(element: HTMLElement) {
-  // Each line triggers a reflow
   element.style.width = '100px'
+  const width = element.offsetWidth  // Forces reflow
   element.style.height = '200px'
-  element.style.backgroundColor = 'blue'
-  element.style.border = '1px solid black'
+  const height = element.offsetHeight  // Forces another reflow
 }
 ```
 
-**Correct: add class - single reflow**
+**Correct: batch writes, then read once**
 
 ```typescript
-// CSS file
-.highlighted-box {
-  width: 100px;
-  height: 200px;
-  background-color: blue;
-  border: 1px solid black;
-}
-
-// JavaScript
 function updateElementStyles(element: HTMLElement) {
   element.classList.add('highlighted-box')
+
+  const { width, height } = element.getBoundingClientRect()
 }
 ```
 
-**Correct: change cssText - single reflow**
+**Better: use CSS classes**
 
-```typescript
-function updateElementStyles(element: HTMLElement) {
-  element.style.cssText = `
-    width: 100px;
-    height: 200px;
-    background-color: blue;
-    border: 1px solid black;
-  `
-}
-```
-
-**React example:**
-
-```tsx
-// Incorrect: changing styles one by one
-function Box({ isHighlighted }: { isHighlighted: boolean }) {
-  const ref = useRef<HTMLDivElement>(null)
-  
-  useEffect(() => {
-    if (ref.current && isHighlighted) {
-      ref.current.style.width = '100px'
-      ref.current.style.height = '200px'
-      ref.current.style.backgroundColor = 'blue'
-    }
-  }, [isHighlighted])
-  
-  return <div ref={ref}>Content</div>
-}
-
-// Correct: toggle class
-function Box({ isHighlighted }: { isHighlighted: boolean }) {
-  return (
-    <div className={isHighlighted ? 'highlighted-box' : ''}>
-      Content
-    </div>
-  )
-}
-```
-
-Prefer CSS classes over inline styles when possible. Classes are cached by the browser and provide better separation of concerns.
+Prefer CSS classes over inline styles when possible. CSS files are cached by the browser, and classes provide better separation of concerns and are easier to maintain.
 
 ### 7.2 Build Index Maps for Repeated Lookups
 
@@ -2044,7 +2150,7 @@ function hasChanges(current: string[], original: string[]) {
   if (current.length !== original.length) {
     return true
   }
-  // Only sort/join when lengths match
+  // Only sort when lengths match
   const currentSorted = current.toSorted()
   const originalSorted = original.toSorted()
   for (let i = 0; i < currentSorted.length; i++) {
@@ -2229,7 +2335,7 @@ const min = Math.min(...numbers)
 const max = Math.max(...numbers)
 ```
 
-This works for small arrays but can be slower for very large arrays due to spread operator limitations. Use the loop approach for reliability.
+This works for small arrays, but can be slower or just throw an error for very large arrays due to spread operator limitations. Maximal array length is approximately 124000 in Chrome 143 and 638000 in Safari 18; exact numbers may vary - see [the fiddle](https://jsfiddle.net/qw1jabsx/4/). Use the loop approach for reliability.
 
 ### 7.11 Use Set/Map for O(1) Lookups
 
@@ -2325,7 +2431,7 @@ Store callbacks in refs when used in effects that shouldn't re-subscribe on call
 **Incorrect: re-subscribes on every render**
 
 ```tsx
-function useWindowEvent(event: string, handler: () => void) {
+function useWindowEvent(event: string, handler: (e) => void) {
   useEffect(() => {
     window.addEventListener(event, handler)
     return () => window.removeEventListener(event, handler)
@@ -2338,7 +2444,7 @@ function useWindowEvent(event: string, handler: () => void) {
 ```tsx
 import { useEffectEvent } from 'react'
 
-function useWindowEvent(event: string, handler: () => void) {
+function useWindowEvent(event: string, handler: (e) => void) {
   const onEvent = useEffectEvent(handler)
 
   useEffect(() => {
@@ -2363,7 +2469,7 @@ Access latest values in callbacks without adding them to dependency arrays. Prev
 ```typescript
 function useLatest<T>(value: T) {
   const ref = useRef(value)
-  useEffect(() => {
+  useLayoutEffect(() => {
     ref.current = value
   }, [value])
   return ref

.agents/skills/vercel-react-best-practices/rules/advanced-use-latest.md

@@ -1,26 +1,14 @@
 ---
-title: useLatest for Stable Callback Refs
+title: useEffectEvent for Stable Callback Refs
 impact: LOW
 impactDescription: prevents effect re-runs
-tags: advanced, hooks, useLatest, refs, optimization
+tags: advanced, hooks, useEffectEvent, refs, optimization
 ---
 
-## useLatest for Stable Callback Refs
+## useEffectEvent for Stable Callback Refs
 
 Access latest values in callbacks without adding them to dependency arrays. Prevents effect re-runs while avoiding stale closures.
 
-**Implementation:**
-
-```typescript
-function useLatest<T>(value: T) {
-  const ref = useRef(value)
-  useLayoutEffect(() => {
-    ref.current = value
-  }, [value])
-  return ref
-}
-```
-
 **Incorrect (effect re-runs on every callback change):**
 
 ```tsx
@@ -34,15 +22,17 @@ function SearchInput({ onSearch }: { onSearch: (q: string) => void }) {
 }
 ```
 
-**Correct (stable effect, fresh callback):**
+**Correct (using React's useEffectEvent):**
 
 ```tsx
+import { useEffectEvent } from 'react';
+
 function SearchInput({ onSearch }: { onSearch: (q: string) => void }) {
   const [query, setQuery] = useState('')
-  const onSearchRef = useLatest(onSearch)
+  const onSearchEvent = useEffectEvent(onSearch)
 
   useEffect(() => {
-    const timeout = setTimeout(() => onSearchRef.current(query), 300)
+    const timeout = setTimeout(() => onSearchEvent(query), 300)
     return () => clearTimeout(timeout)
   }, [query])
 }

.agents/skills/vercel-react-best-practices/rules/async-dependencies.md

@@ -33,4 +33,19 @@ const { user, config, profile } = await all({
 })
 ```
 
+**Alternative without extra dependencies:**
+
+We can also create all the promises first, and do `Promise.all()` at the end.
+
+```typescript
+const userPromise = fetchUser()
+const profilePromise = userPromise.then(user => fetchProfile(user.id))
+
+const [user, config, profile] = await Promise.all([
+  userPromise,
+  fetchConfig(),
+  profilePromise
+])
+```
+
 Reference: [https://github.com/shuding/better-all](https://github.com/shuding/better-all)

.agents/skills/vercel-react-best-practices/rules/js-batch-dom-css.md

@@ -1,18 +1,28 @@
 ---
-title: Batch DOM CSS Changes
+title: Avoid Layout Thrashing
 impact: MEDIUM
-impactDescription: reduces reflows/repaints
-tags: javascript, dom, css, performance, reflow
+impactDescription: prevents forced synchronous layouts and reduces performance bottlenecks
+tags: javascript, dom, css, performance, reflow, layout-thrashing
 ---
 
-## Batch DOM CSS Changes
+## Avoid Layout Thrashing
 
 Avoid interleaving style writes with layout reads. When you read a layout property (like `offsetWidth`, `getBoundingClientRect()`, or `getComputedStyle()`) between style changes, the browser is forced to trigger a synchronous reflow.
 
-**Incorrect (interleaved reads and writes force reflows):**
-
+**This is OK (browser batches style changes):**
 ```typescript
 function updateElementStyles(element: HTMLElement) {
+  // Each line invalidates style, but browser batches the recalculation
+  element.style.width = '100px'
+  element.style.height = '200px'
+  element.style.backgroundColor = 'blue'
+  element.style.border = '1px solid black'
+}
+```
+
+**Incorrect (interleaved reads and writes force reflows):**
+```typescript
+function layoutThrashing(element: HTMLElement) {
   element.style.width = '100px'
   const width = element.offsetWidth  // Forces reflow
   element.style.height = '200px'
@@ -21,7 +31,6 @@ function updateElementStyles(element: HTMLElement) {
 ```
 
 **Correct (batch writes, then read once):**
-
 ```typescript
 function updateElementStyles(element: HTMLElement) {
   // Batch all writes together
@@ -35,8 +44,21 @@ function updateElementStyles(element: HTMLElement) {
 }
 ```
 
-**Better: use CSS classes**
+**Correct (batch reads, then writes):**
+```typescript
+function avoidThrashing(element: HTMLElement) {
+  // Read phase - all layout queries first
+  const rect1 = element.getBoundingClientRect()
+  const offsetWidth = element.offsetWidth
+  const offsetHeight = element.offsetHeight
+  
+  // Write phase - all style changes after
+  element.style.width = '100px'
+  element.style.height = '200px'
+}
+```
 
+**Better: use CSS classes**
 ```css
 .highlighted-box {
   width: 100px;
@@ -45,13 +67,41 @@ function updateElementStyles(element: HTMLElement) {
   border: 1px solid black;
 }
 ```
-
 ```typescript
 function updateElementStyles(element: HTMLElement) {
   element.classList.add('highlighted-box')
-
+  
   const { width, height } = element.getBoundingClientRect()
 }
 ```
 
-Prefer CSS classes over inline styles when possible. CSS files are cached by the browser, and classes provide better separation of concerns and are easier to maintain.
+**React example:**
+```tsx
+// Incorrect: interleaving style changes with layout queries
+function Box({ isHighlighted }: { isHighlighted: boolean }) {
+  const ref = useRef<HTMLDivElement>(null)
+  
+  useEffect(() => {
+    if (ref.current && isHighlighted) {
+      ref.current.style.width = '100px'
+      const width = ref.current.offsetWidth // Forces layout
+      ref.current.style.height = '200px'
+    }
+  }, [isHighlighted])
+  
+  return <div ref={ref}>Content</div>
+}
+
+// Correct: toggle class
+function Box({ isHighlighted }: { isHighlighted: boolean }) {
+  return (
+    <div className={isHighlighted ? 'highlighted-box' : ''}>
+      Content
+    </div>
+  )
+}
+```
+
+Prefer CSS classes over inline styles when possible. CSS files are cached by the browser, and classes provide better separation of concerns and are easier to maintain.
+
+See [this gist](https://gist.github.com/paulirish/5d52fb081b3570c81e3a) and [CSS Triggers](https://csstriggers.com/) for more information on layout-forcing operations.

.agents/skills/vercel-react-best-practices/rules/rerender-memo-with-default-value.md

@@ -0,0 +1,38 @@
+---
+
+title: Extract Default Non-primitive Parameter Value from Memoized Component to Constant
+impact: MEDIUM
+impactDescription: restores memoization by using a constant for default value
+tags: rerender, memo, optimization
+
+---
+
+## Extract Default Non-primitive Parameter Value from Memoized Component to Constant
+
+When memoized component has a default value for some non-primitive optional parameter, such as an array, function, or object, calling the component without that parameter results in broken memoization. This is because new value instances are created on every rerender, and they do not pass strict equality comparison in `memo()`.
+
+To address this issue, extract the default value into a constant.
+
+**Incorrect (`onClick` has different values on every rerender):**
+
+```tsx
+const UserAvatar = memo(function UserAvatar({ onClick = () => {} }: { onClick?: () => void }) {
+  // ...
+})
+
+// Used without optional onClick
+<UserAvatar />
+```
+
+**Correct (stable default value):**
+
+```tsx
+const NOOP = () => {};
+
+const UserAvatar = memo(function UserAvatar({ onClick = NOOP }: { onClick?: () => void }) {
+  // ...
+})
+
+// Used without optional onClick
+<UserAvatar />
+```

.agents/skills/vercel-react-best-practices/rules/rerender-simple-expression-in-memo.md

@@ -0,0 +1,35 @@
+---
+title: Do not wrap a simple expression with a primitive result type in useMemo
+impact: LOW-MEDIUM
+impactDescription: wasted computation on every render
+tags: rerender, useMemo, optimization
+---
+
+## Do not wrap a simple expression with a primitive result type in useMemo
+
+When an expression is simple (few logical or arithmetical operators) and has a primitive result type (boolean, number, string), do not wrap it in `useMemo`.
+Calling `useMemo` and comparing hook dependencies may consume more resources than the expression itself.
+
+**Incorrect:**
+
+```tsx
+function Header({ user, notifications }: Props) {
+  const isLoading = useMemo(() => {
+    return user.isLoading || notifications.isLoading
+  }, [user.isLoading, notifications.isLoading])
+
+  if (isLoading) return <Skeleton />
+  // return some markup
+}
+```
+
+**Correct:**
+
+```tsx
+function Header({ user, notifications }: Props) {
+  const isLoading = user.isLoading || notifications.isLoading
+
+  if (isLoading) return <Skeleton />
+  // return some markup
+}
+```

.agents/skills/vercel-react-best-practices/rules/server-auth-actions.md

@@ -0,0 +1,96 @@
+---
+title: Authenticate Server Actions Like API Routes
+impact: CRITICAL
+impactDescription: prevents unauthorized access to server mutations
+tags: server, server-actions, authentication, security, authorization
+---
+
+## Authenticate Server Actions Like API Routes
+
+**Impact: CRITICAL (prevents unauthorized access to server mutations)**
+
+Server Actions (functions with `"use server"`) are exposed as public endpoints, just like API routes. Always verify authentication and authorization **inside** each Server Action—do not rely solely on middleware, layout guards, or page-level checks, as Server Actions can be invoked directly.
+
+Next.js documentation explicitly states: "Treat Server Actions with the same security considerations as public-facing API endpoints, and verify if the user is allowed to perform a mutation."
+
+**Incorrect (no authentication check):**
+
+```typescript
+'use server'
+
+export async function deleteUser(userId: string) {
+  // Anyone can call this! No auth check
+  await db.user.delete({ where: { id: userId } })
+  return { success: true }
+}
+```
+
+**Correct (authentication inside the action):**
+
+```typescript
+'use server'
+
+import { verifySession } from '@/lib/auth'
+import { unauthorized } from '@/lib/errors'
+
+export async function deleteUser(userId: string) {
+  // Always check auth inside the action
+  const session = await verifySession()
+  
+  if (!session) {
+    throw unauthorized('Must be logged in')
+  }
+  
+  // Check authorization too
+  if (session.user.role !== 'admin' && session.user.id !== userId) {
+    throw unauthorized('Cannot delete other users')
+  }
+  
+  await db.user.delete({ where: { id: userId } })
+  return { success: true }
+}
+```
+
+**With input validation:**
+
+```typescript
+'use server'
+
+import { verifySession } from '@/lib/auth'
+import { z } from 'zod'
+
+const updateProfileSchema = z.object({
+  userId: z.string().uuid(),
+  name: z.string().min(1).max(100),
+  email: z.string().email()
+})
+
+export async function updateProfile(data: unknown) {
+  // Validate input first
+  const validated = updateProfileSchema.parse(data)
+  
+  // Then authenticate
+  const session = await verifySession()
+  if (!session) {
+    throw new Error('Unauthorized')
+  }
+  
+  // Then authorize
+  if (session.user.id !== validated.userId) {
+    throw new Error('Can only update own profile')
+  }
+  
+  // Finally perform the mutation
+  await db.user.update({
+    where: { id: validated.userId },
+    data: {
+      name: validated.name,
+      email: validated.email
+    }
+  })
+  
+  return { success: true }
+}
+```
+
+Reference: [https://nextjs.org/docs/app/guides/authentication](https://nextjs.org/docs/app/guides/authentication)

.agents/skills/vercel-react-best-practices/rules/server-dedup-props.md

@@ -0,0 +1,65 @@
+---
+title: Avoid Duplicate Serialization in RSC Props
+impact: LOW
+impactDescription: reduces network payload by avoiding duplicate serialization
+tags: server, rsc, serialization, props, client-components
+---
+
+## Avoid Duplicate Serialization in RSC Props
+
+**Impact: LOW (reduces network payload by avoiding duplicate serialization)**
+
+RSC→client serialization deduplicates by object reference, not value. Same reference = serialized once; new reference = serialized again. Do transformations (`.toSorted()`, `.filter()`, `.map()`) in client, not server.
+
+**Incorrect (duplicates array):**
+
+```tsx
+// RSC: sends 6 strings (2 arrays × 3 items)
+<ClientList usernames={usernames} usernamesOrdered={usernames.toSorted()} />
+```
+
+**Correct (sends 3 strings):**
+
+```tsx
+// RSC: send once
+<ClientList usernames={usernames} />
+
+// Client: transform there
+'use client'
+const sorted = useMemo(() => [...usernames].sort(), [usernames])
+```
+
+**Nested deduplication behavior:**
+
+Deduplication works recursively. Impact varies by data type:
+
+- `string[]`, `number[]`, `boolean[]`: **HIGH impact** - array + all primitives fully duplicated
+- `object[]`: **LOW impact** - array duplicated, but nested objects deduplicated by reference
+
+```tsx
+// string[] - duplicates everything
+usernames={['a','b']} sorted={usernames.toSorted()} // sends 4 strings
+
+// object[] - duplicates array structure only
+users={[{id:1},{id:2}]} sorted={users.toSorted()} // sends 2 arrays + 2 unique objects (not 4)
+```
+
+**Operations breaking deduplication (create new references):**
+
+- Arrays: `.toSorted()`, `.filter()`, `.map()`, `.slice()`, `[...arr]`
+- Objects: `{...obj}`, `Object.assign()`, `structuredClone()`, `JSON.parse(JSON.stringify())`
+
+**More examples:**
+
+```tsx
+// ❌ Bad
+<C users={users} active={users.filter(u => u.active)} />
+<C product={product} productName={product.name} />
+
+// ✅ Good
+<C users={users} />
+<C product={product} />
+// Do filtering/destructuring in client
+```
+
+**Exception:** Pass derived data when transformation is expensive or client doesn't need original.

.gitignore

@@ -209,6 +209,7 @@ api/.vscode
 .history
 
 .idea/
+web/migration/
 
 # pnpm
 /.pnpm-store

api/.env.example

@@ -717,3 +717,13 @@ SANDBOX_EXPIRED_RECORDS_CLEAN_BATCH_SIZE=1000
 SANDBOX_EXPIRED_RECORDS_RETENTION_DAYS=30
 SANDBOX_EXPIRED_RECORDS_CLEAN_TASK_LOCK_TTL=90000
 
+# Sandbox Dify CLI configuration
+# Directory containing dify CLI binaries (dify-cli-<os>-<arch>). Defaults to api/bin when unset.
+SANDBOX_DIFY_CLI_ROOT=
+
+# CLI API URL for sandbox (dify-sandbox or e2b) to call back to Dify API.
+# This URL must be accessible from the sandbox environment.
+# For local development: use http://localhost:5001 or http://127.0.0.1:5001
+# For Docker deployment: use http://api:5001 (internal Docker network)
+# For external sandbox (e.g., e2b): use a publicly accessible URL
+CLI_API_URL=http://localhost:5001

api/agent-notes/core/app_assets/packager/zip_packager.py.md

@@ -0,0 +1,14 @@
+# Zip Packager Notes
+
+## Purpose
+- Builds a ZIP archive of asset contents stored via the configured storage backend.
+
+## Key Decisions
+- Packaging writes assets into an in-memory zip buffer returned as bytes.
+- Asset fetch + zip writing are executed via a thread pool with a lock guarding `ZipFile` writes.
+
+## Edge Cases
+- ZIP writes are serialized by the lock; storage reads still run in parallel.
+
+## Tests/Verification
+- None yet.

api/agent-notes/core/virtual_environment/providers/e2b_sandbox.py.md

@@ -0,0 +1,16 @@
+# E2B Sandbox Provider Notes
+
+## Purpose
+- Implements the E2B-backed `VirtualEnvironment` provider and bootstraps sandbox metadata, file I/O, and command execution.
+
+## Key Decisions
+- Sandbox metadata is gathered during `_construct_environment` using the E2B SDK before returning `Metadata`.
+- Architecture/OS detection uses a single `uname -m -s` call split by whitespace to reduce round-trips.
+- Command execution streams stdout/stderr through `QueueTransportReadCloser`; stdin is unsupported.
+
+## Edge Cases
+- `release_environment` raises when sandbox termination fails.
+- `execute_command` runs in a background thread; consumers must read stdout/stderr until EOF.
+
+## Tests/Verification
+- None yet. Add targeted service tests when behavior changes.

api/commands.py

@@ -23,7 +23,8 @@ from core.rag.datasource.vdb.vector_factory import Vector
 from core.rag.datasource.vdb.vector_type import VectorType
 from core.rag.index_processor.constant.built_in_field import BuiltInField
 from core.rag.models.document import Document
-from core.tools.utils.system_oauth_encryption import encrypt_system_oauth_params
+from core.sandbox import SandboxBuilder, SandboxType
+from core.tools.utils.system_encryption import encrypt_system_params
 from events.app_event import app_was_created
 from extensions.ext_database import db
 from extensions.ext_redis import redis_client
@@ -1245,7 +1246,7 @@ def remove_orphaned_files_on_storage(force: bool):
             click.echo(click.style(f"- Scanning files on storage path {storage_path}", fg="white"))
             files = storage.scan(path=storage_path, files=True, directories=False)
             all_files_on_storage.extend(files)
-        except FileNotFoundError as e:
+        except FileNotFoundError:
             click.echo(click.style(f"  -> Skipping path {storage_path} as it does not exist.", fg="yellow"))
             continue
         except Exception as e:
@@ -1493,6 +1494,57 @@ def file_usage(
             click.echo(click.style(f"Use --offset {offset + limit} to see next page", fg="white"))
 
 
+@click.command("setup-sandbox-system-config", help="Setup system-level sandbox provider configuration.")
+@click.option(
+    "--provider-type", prompt=True, type=click.Choice(["e2b", "docker", "local"]), help="Sandbox provider type"
+)
+@click.option("--config", prompt=True, help='Configuration JSON (e.g., {"api_key": "xxx"} for e2b)')
+def setup_sandbox_system_config(provider_type: str, config: str):
+    """
+    Setup system-level sandbox provider configuration.
+
+    Examples:
+        flask setup-sandbox-system-config --provider-type e2b --config '{"api_key": "e2b_xxx"}'
+        flask setup-sandbox-system-config --provider-type docker --config '{"docker_sock": "unix:///var/run/docker.sock"}'
+        flask setup-sandbox-system-config --provider-type local --config '{}'
+    """
+    from models.sandbox import SandboxProviderSystemConfig
+
+    try:
+        click.echo(click.style(f"Validating config: {config}", fg="yellow"))
+        config_dict = TypeAdapter(dict[str, Any]).validate_json(config)
+        click.echo(click.style("Config validated successfully.", fg="green"))
+
+        click.echo(click.style(f"Validating config schema for provider type: {provider_type}", fg="yellow"))
+        SandboxBuilder.validate(SandboxType(provider_type), config_dict)
+        click.echo(click.style("Config schema validated successfully.", fg="green"))
+
+        click.echo(click.style("Encrypting config...", fg="yellow"))
+        click.echo(click.style(f"Using SECRET_KEY: `{dify_config.SECRET_KEY}`", fg="yellow"))
+        encrypted_config = encrypt_system_params(config_dict)
+        click.echo(click.style("Config encrypted successfully.", fg="green"))
+    except Exception as e:
+        click.echo(click.style(f"Error validating/encrypting config: {str(e)}", fg="red"))
+        return
+
+    deleted_count = db.session.query(SandboxProviderSystemConfig).filter_by(provider_type=provider_type).delete()
+    if deleted_count > 0:
+        click.echo(
+            click.style(
+                f"Deleted {deleted_count} existing system config for provider type: {provider_type}", fg="yellow"
+            )
+        )
+
+    system_config = SandboxProviderSystemConfig(
+        provider_type=provider_type,
+        encrypted_config=encrypted_config,
+    )
+    db.session.add(system_config)
+    db.session.commit()
+    click.echo(click.style(f"Sandbox system config setup successfully. id: {system_config.id}", fg="green"))
+    click.echo(click.style(f"Provider type: {provider_type}", fg="green"))
+
+
 @click.command("setup-system-tool-oauth-client", help="Setup system tool oauth client.")
 @click.option("--provider", prompt=True, help="Provider name")
 @click.option("--client-params", prompt=True, help="Client Params")
@@ -1512,7 +1564,7 @@ def setup_system_tool_oauth_client(provider, client_params):
 
         click.echo(click.style(f"Encrypting client params: {client_params}", fg="yellow"))
         click.echo(click.style(f"Using SECRET_KEY: `{dify_config.SECRET_KEY}`", fg="yellow"))
-        oauth_client_params = encrypt_system_oauth_params(client_params_dict)
+        oauth_client_params = encrypt_system_params(client_params_dict)
         click.echo(click.style("Client params encrypted successfully.", fg="green"))
     except Exception as e:
         click.echo(click.style(f"Error parsing client params: {str(e)}", fg="red"))
@@ -1561,7 +1613,7 @@ def setup_system_trigger_oauth_client(provider, client_params):
 
         click.echo(click.style(f"Encrypting client params: {client_params}", fg="yellow"))
         click.echo(click.style(f"Using SECRET_KEY: `{dify_config.SECRET_KEY}`", fg="yellow"))
-        oauth_client_params = encrypt_system_oauth_params(client_params_dict)
+        oauth_client_params = encrypt_system_params(client_params_dict)
         click.echo(click.style("Client params encrypted successfully.", fg="green"))
     except Exception as e:
         click.echo(click.style(f"Error parsing client params: {str(e)}", fg="red"))

api/configs/app_config.py

@@ -2,6 +2,7 @@ import logging
 from pathlib import Path
 from typing import Any
 
+from pydantic import Field
 from pydantic.fields import FieldInfo
 from pydantic_settings import BaseSettings, PydanticBaseSettingsSource, SettingsConfigDict, TomlConfigSettingsSource
 
@@ -82,6 +83,14 @@ class DifyConfig(
         extra="ignore",
     )
 
+    SANDBOX_DIFY_CLI_ROOT: str | None = Field(
+        default=None,
+        description=(
+            "Filesystem directory containing dify CLI binaries named dify-cli-<os>-<arch>. "
+            "Defaults to api/bin when unset."
+        ),
+    )
+
     # Before adding any config,
     # please consider to arrange it in the proper config group of existed or added
     # for better readability and maintainability.

api/configs/feature/__init__.py

@@ -244,6 +244,17 @@ class PluginConfig(BaseSettings):
     )
 
 
+class CliApiConfig(BaseSettings):
+    """
+    Configuration for CLI API (for dify-cli to call back from external sandbox environments)
+    """
+
+    CLI_API_URL: str = Field(
+        description="CLI API URL for external sandbox (e.g., e2b) to call back.",
+        default="http://localhost:5001",
+    )
+
+
 class MarketplaceConfig(BaseSettings):
     """
     Configuration for marketplace
@@ -1323,6 +1334,7 @@ class FeatureConfig(
     TriggerConfig,
     AsyncWorkflowConfig,
     PluginConfig,
+    CliApiConfig,
     MarketplaceConfig,
     DataSetConfig,
     EndpointConfig,

api/controllers/cli_api/__init__.py

@@ -0,0 +1,27 @@
+from flask import Blueprint
+from flask_restx import Namespace
+
+from libs.external_api import ExternalApi
+
+bp = Blueprint("cli_api", __name__, url_prefix="/cli/api")
+
+api = ExternalApi(
+    bp,
+    version="1.0",
+    title="CLI API",
+    description="APIs for Dify CLI to call back from external sandbox environments (e.g., e2b)",
+)
+
+# Create namespace
+cli_api_ns = Namespace("cli_api", description="CLI API operations", path="/")
+
+from .plugin import plugin as _plugin
+
+api.add_namespace(cli_api_ns)
+
+__all__ = [
+    "_plugin",
+    "api",
+    "bp",
+    "cli_api_ns",
+]

api/controllers/cli_api/plugin/plugin.py

@@ -0,0 +1,137 @@
+from flask_restx import Resource
+
+from controllers.cli_api import cli_api_ns
+from controllers.cli_api.plugin.wraps import get_cli_user_tenant, plugin_data
+from controllers.cli_api.wraps import cli_api_only
+from controllers.console.wraps import setup_required
+from core.file.helpers import get_signed_file_url_for_plugin
+from core.plugin.backwards_invocation.app import PluginAppBackwardsInvocation
+from core.plugin.backwards_invocation.base import BaseBackwardsInvocationResponse
+from core.plugin.backwards_invocation.model import PluginModelBackwardsInvocation
+from core.plugin.backwards_invocation.tool import PluginToolBackwardsInvocation
+from core.plugin.entities.request import (
+    RequestInvokeApp,
+    RequestInvokeLLM,
+    RequestInvokeTool,
+    RequestRequestUploadFile,
+)
+from core.tools.entities.tool_entities import ToolProviderType
+from libs.helper import length_prefixed_response
+from models import Account, Tenant
+from models.model import EndUser
+
+
+@cli_api_ns.route("/invoke/llm")
+class CliInvokeLLMApi(Resource):
+    @get_cli_user_tenant
+    @setup_required
+    @cli_api_only
+    @plugin_data(payload_type=RequestInvokeLLM)
+    def post(self, user_model: Account | EndUser, tenant_model: Tenant, payload: RequestInvokeLLM):
+        def generator():
+            response = PluginModelBackwardsInvocation.invoke_llm(user_model.id, tenant_model, payload)
+            return PluginModelBackwardsInvocation.convert_to_event_stream(response)
+
+        return length_prefixed_response(0xF, generator())
+
+
+@cli_api_ns.route("/invoke/tool")
+class CliInvokeToolApi(Resource):
+    @get_cli_user_tenant
+    @setup_required
+    @cli_api_only
+    @plugin_data(payload_type=RequestInvokeTool)
+    def post(self, user_model: Account | EndUser, tenant_model: Tenant, payload: RequestInvokeTool):
+        def generator():
+            return PluginToolBackwardsInvocation.convert_to_event_stream(
+                PluginToolBackwardsInvocation.invoke_tool(
+                    tenant_id=tenant_model.id,
+                    user_id=user_model.id,
+                    tool_type=ToolProviderType.value_of(payload.tool_type),
+                    provider=payload.provider,
+                    tool_name=payload.tool,
+                    tool_parameters=payload.tool_parameters,
+                    credential_id=payload.credential_id,
+                ),
+            )
+
+        return length_prefixed_response(0xF, generator())
+
+
+@cli_api_ns.route("/invoke/app")
+class CliInvokeAppApi(Resource):
+    @get_cli_user_tenant
+    @setup_required
+    @cli_api_only
+    @plugin_data(payload_type=RequestInvokeApp)
+    def post(self, user_model: Account | EndUser, tenant_model: Tenant, payload: RequestInvokeApp):
+        response = PluginAppBackwardsInvocation.invoke_app(
+            app_id=payload.app_id,
+            user_id=user_model.id,
+            tenant_id=tenant_model.id,
+            conversation_id=payload.conversation_id,
+            query=payload.query,
+            stream=payload.response_mode == "streaming",
+            inputs=payload.inputs,
+            files=payload.files,
+        )
+
+        return length_prefixed_response(0xF, PluginAppBackwardsInvocation.convert_to_event_stream(response))
+
+
+@cli_api_ns.route("/upload/file/request")
+class CliUploadFileRequestApi(Resource):
+    @get_cli_user_tenant
+    @setup_required
+    @cli_api_only
+    @plugin_data(payload_type=RequestRequestUploadFile)
+    def post(self, user_model: Account | EndUser, tenant_model: Tenant, payload: RequestRequestUploadFile):
+        # generate signed url
+        url = get_signed_file_url_for_plugin(
+            filename=payload.filename,
+            mimetype=payload.mimetype,
+            tenant_id=tenant_model.id,
+            user_id=user_model.id,
+        )
+        return BaseBackwardsInvocationResponse(data={"url": url}).model_dump()
+
+
+@cli_api_ns.route("/fetch/tools/list")
+class CliFetchToolsListApi(Resource):
+    @get_cli_user_tenant
+    @setup_required
+    @cli_api_only
+    def post(self, user_model: Account | EndUser, tenant_model: Tenant):
+        from sqlalchemy.orm import Session
+
+        from extensions.ext_database import db
+        from services.tools.api_tools_manage_service import ApiToolManageService
+        from services.tools.builtin_tools_manage_service import BuiltinToolManageService
+        from services.tools.mcp_tools_manage_service import MCPToolManageService
+        from services.tools.workflow_tools_manage_service import WorkflowToolManageService
+
+        providers = []
+
+        # Get builtin tools
+        builtin_providers = BuiltinToolManageService.list_builtin_tools(user_model.id, tenant_model.id)
+        for provider in builtin_providers:
+            providers.append(provider.to_dict())
+
+        # Get API tools
+        api_providers = ApiToolManageService.list_api_tools(tenant_model.id)
+        for provider in api_providers:
+            providers.append(provider.to_dict())
+
+        # Get workflow tools
+        workflow_providers = WorkflowToolManageService.list_tenant_workflow_tools(user_model.id, tenant_model.id)
+        for provider in workflow_providers:
+            providers.append(provider.to_dict())
+
+        # Get MCP tools
+        with Session(db.engine) as session:
+            mcp_service = MCPToolManageService(session)
+            mcp_providers = mcp_service.list_providers(tenant_id=tenant_model.id, for_list=True)
+            for provider in mcp_providers:
+                providers.append(provider.to_dict())
+
+        return BaseBackwardsInvocationResponse(data={"providers": providers}).model_dump()

api/controllers/cli_api/plugin/wraps.py

@@ -0,0 +1,146 @@
+from collections.abc import Callable
+from functools import wraps
+from typing import ParamSpec, TypeVar
+
+from flask import current_app, request
+from flask_login import user_logged_in
+from pydantic import BaseModel
+from sqlalchemy.orm import Session
+
+from core.session.cli_api import CliApiSession, CliApiSessionManager
+from extensions.ext_database import db
+from libs.login import current_user
+from models.account import Tenant
+from models.model import DefaultEndUserSessionID, EndUser
+
+P = ParamSpec("P")
+R = TypeVar("R")
+
+
+class TenantUserPayload(BaseModel):
+    tenant_id: str
+    user_id: str
+
+
+def get_user(tenant_id: str, user_id: str | None) -> EndUser:
+    """
+    Get current user
+
+    NOTE: user_id is not trusted, it could be maliciously set to any value.
+    As a result, it could only be considered as an end user id.
+    """
+    if not user_id:
+        user_id = DefaultEndUserSessionID.DEFAULT_SESSION_ID
+    is_anonymous = user_id == DefaultEndUserSessionID.DEFAULT_SESSION_ID
+    try:
+        with Session(db.engine) as session:
+            user_model = None
+
+            if is_anonymous:
+                user_model = (
+                    session.query(EndUser)
+                    .where(
+                        EndUser.session_id == user_id,
+                        EndUser.tenant_id == tenant_id,
+                    )
+                    .first()
+                )
+            else:
+                user_model = (
+                    session.query(EndUser)
+                    .where(
+                        EndUser.id == user_id,
+                        EndUser.tenant_id == tenant_id,
+                    )
+                    .first()
+                )
+
+            if not user_model:
+                user_model = EndUser(
+                    tenant_id=tenant_id,
+                    type="service_api",
+                    is_anonymous=is_anonymous,
+                    session_id=user_id,
+                )
+                session.add(user_model)
+                session.commit()
+                session.refresh(user_model)
+
+    except Exception:
+        raise ValueError("user not found")
+
+    return user_model
+
+
+def get_cli_user_tenant(view_func: Callable[P, R]):
+    @wraps(view_func)
+    def decorated_view(*args: P.args, **kwargs: P.kwargs):
+        session_id = request.headers.get("X-Cli-Api-Session-Id")
+
+        if session_id:
+            session: CliApiSession | None = CliApiSessionManager().get(session_id)
+            if not session:
+                raise ValueError("session not found")
+            user_id = session.user_id
+            tenant_id = session.tenant_id
+
+        else:
+            payload = TenantUserPayload.model_validate(request.get_json(silent=True) or {})
+            user_id = payload.user_id
+            tenant_id = payload.tenant_id
+
+        if not tenant_id:
+            raise ValueError("tenant_id is required")
+
+        if not user_id:
+            user_id = DefaultEndUserSessionID.DEFAULT_SESSION_ID
+
+        try:
+            tenant_model = (
+                db.session.query(Tenant)
+                .where(
+                    Tenant.id == tenant_id,
+                )
+                .first()
+            )
+        except Exception:
+            raise ValueError("tenant not found")
+
+        if not tenant_model:
+            raise ValueError("tenant not found")
+
+        kwargs["tenant_model"] = tenant_model
+
+        user = get_user(tenant_id, user_id)
+        kwargs["user_model"] = user
+
+        current_app.login_manager._update_request_context_with_user(user)  # type: ignore
+        user_logged_in.send(current_app._get_current_object(), user=current_user)  # type: ignore
+
+        return view_func(*args, **kwargs)
+
+    return decorated_view
+
+
+def plugin_data(view: Callable[P, R] | None = None, *, payload_type: type[BaseModel]):
+    def decorator(view_func: Callable[P, R]):
+        def decorated_view(*args: P.args, **kwargs: P.kwargs):
+            try:
+                data = request.get_json()
+            except Exception:
+                raise ValueError("invalid json")
+
+            try:
+                payload = payload_type.model_validate(data)
+            except Exception as e:
+                raise ValueError(f"invalid payload: {str(e)}")
+
+            kwargs["payload"] = payload
+            return view_func(*args, **kwargs)
+
+        return decorated_view
+
+    if view is None:
+        return decorator
+    else:
+        return decorator(view)

api/controllers/cli_api/wraps.py

@@ -0,0 +1,54 @@
+import hashlib
+import hmac
+import time
+from collections.abc import Callable
+from functools import wraps
+from typing import ParamSpec, TypeVar
+
+from flask import abort, request
+
+from core.session.cli_api import CliApiSessionManager
+
+P = ParamSpec("P")
+R = TypeVar("R")
+
+SIGNATURE_TTL_SECONDS = 300
+
+
+def _verify_signature(session_secret: str, timestamp: str, body: bytes, signature: str) -> bool:
+    expected = hmac.new(
+        session_secret.encode(),
+        f"{timestamp}.".encode() + body,
+        hashlib.sha256,
+    ).hexdigest()
+    return hmac.compare_digest(f"sha256={expected}", signature)
+
+
+def cli_api_only(view: Callable[P, R]):
+    @wraps(view)
+    def decorated(*args: P.args, **kwargs: P.kwargs):
+        session_id = request.headers.get("X-Cli-Api-Session-Id")
+        timestamp = request.headers.get("X-Cli-Api-Timestamp")
+        signature = request.headers.get("X-Cli-Api-Signature")
+
+        if not session_id or not timestamp or not signature:
+            abort(401)
+
+        try:
+            ts = int(timestamp)
+            if abs(time.time() - ts) > SIGNATURE_TTL_SECONDS:
+                abort(401)
+        except ValueError:
+            abort(401)
+
+        session = CliApiSessionManager().get(session_id)
+        if not session:
+            abort(401)
+
+        body = request.get_data()
+        if not _verify_signature(session.secret, timestamp, body, signature):
+            abort(401)
+
+        return view(*args, **kwargs)
+
+    return decorated

api/controllers/console/__init__.py

@@ -50,6 +50,7 @@ from .app import (
     agent,
     annotation,
     app,
+    app_asset,
     audio,
     completion,
     conversation,
@@ -128,6 +129,7 @@ from .workspace import (
     model_providers,
     models,
     plugin,
+    sandbox_providers,
     tool_providers,
     trigger_providers,
     workspace,
@@ -146,6 +148,7 @@ __all__ = [
     "api",
     "apikey",
     "app",
+    "app_asset",
     "audio",
     "banner",
     "billing",
@@ -194,6 +197,7 @@ __all__ = [
     "rag_pipeline_import",
     "rag_pipeline_workflow",
     "recommended_app",
+    "sandbox_providers",
     "saved_message",
     "setup",
     "site",

api/controllers/console/app/app_asset.py

@@ -0,0 +1,274 @@
+from flask import request
+from flask_restx import Resource
+from pydantic import BaseModel, Field, field_validator
+
+from controllers.console import console_ns
+from controllers.console.app.error import (
+    AppAssetFileRequiredError,
+    AppAssetNodeNotFoundError,
+    AppAssetPathConflictError,
+)
+from controllers.console.app.wraps import get_app_model
+from controllers.console.wraps import account_initialization_required, setup_required
+from libs.login import current_account_with_tenant, login_required
+from models import App
+from models.model import AppMode
+from services.app_asset_service import AppAssetService
+from services.errors.app_asset import (
+    AppAssetNodeNotFoundError as ServiceNodeNotFoundError,
+)
+from services.errors.app_asset import (
+    AppAssetParentNotFoundError,
+)
+from services.errors.app_asset import (
+    AppAssetPathConflictError as ServicePathConflictError,
+)
+
+DEFAULT_REF_TEMPLATE_SWAGGER_2_0 = "#/definitions/{model}"
+
+
+class CreateFolderPayload(BaseModel):
+    name: str = Field(..., min_length=1, max_length=255)
+    parent_id: str | None = None
+
+
+class CreateFilePayload(BaseModel):
+    name: str = Field(..., min_length=1, max_length=255)
+    parent_id: str | None = None
+
+    @field_validator("name", mode="before")
+    @classmethod
+    def strip_name(cls, v: str) -> str:
+        return v.strip() if isinstance(v, str) else v
+
+    @field_validator("parent_id", mode="before")
+    @classmethod
+    def empty_to_none(cls, v: str | None) -> str | None:
+        return v or None
+
+
+class UpdateFileContentPayload(BaseModel):
+    content: str
+
+
+class RenameNodePayload(BaseModel):
+    name: str = Field(..., min_length=1, max_length=255)
+
+
+class MoveNodePayload(BaseModel):
+    parent_id: str | None = None
+
+
+class ReorderNodePayload(BaseModel):
+    after_node_id: str | None = Field(default=None, description="Place after this node, None for first position")
+
+
+def reg(cls: type[BaseModel]) -> None:
+    console_ns.schema_model(cls.__name__, cls.model_json_schema(ref_template=DEFAULT_REF_TEMPLATE_SWAGGER_2_0))
+
+
+reg(CreateFolderPayload)
+reg(CreateFilePayload)
+reg(UpdateFileContentPayload)
+reg(RenameNodePayload)
+reg(MoveNodePayload)
+reg(ReorderNodePayload)
+
+
+@console_ns.route("/apps/<string:app_id>/assets/tree")
+class AppAssetTreeResource(Resource):
+    @setup_required
+    @login_required
+    @account_initialization_required
+    @get_app_model(mode=[AppMode.ADVANCED_CHAT, AppMode.WORKFLOW])
+    def get(self, app_model: App):
+        current_user, _ = current_account_with_tenant()
+        tree = AppAssetService.get_asset_tree(app_model, current_user.id)
+        return {"children": [view.model_dump() for view in tree.transform()]}
+
+
+@console_ns.route("/apps/<string:app_id>/assets/folders")
+class AppAssetFolderResource(Resource):
+    @console_ns.expect(console_ns.models[CreateFolderPayload.__name__])
+    @setup_required
+    @login_required
+    @account_initialization_required
+    @get_app_model(mode=[AppMode.ADVANCED_CHAT, AppMode.WORKFLOW])
+    def post(self, app_model: App):
+        current_user, _ = current_account_with_tenant()
+        payload = CreateFolderPayload.model_validate(console_ns.payload or {})
+
+        try:
+            node = AppAssetService.create_folder(app_model, current_user.id, payload.name, payload.parent_id)
+            return node.model_dump(), 201
+        except AppAssetParentNotFoundError:
+            raise AppAssetNodeNotFoundError()
+        except ServicePathConflictError:
+            raise AppAssetPathConflictError()
+
+
+@console_ns.route("/apps/<string:app_id>/assets/files")
+class AppAssetFileResource(Resource):
+    @setup_required
+    @login_required
+    @account_initialization_required
+    @get_app_model(mode=[AppMode.ADVANCED_CHAT, AppMode.WORKFLOW])
+    def post(self, app_model: App):
+        current_user, _ = current_account_with_tenant()
+
+        file = request.files.get("file")
+        if not file:
+            raise AppAssetFileRequiredError()
+
+        payload = CreateFilePayload.model_validate(request.form.to_dict())
+        content = file.read()
+
+        try:
+            node = AppAssetService.create_file(app_model, current_user.id, payload.name, content, payload.parent_id)
+            return node.model_dump(), 201
+        except AppAssetParentNotFoundError:
+            raise AppAssetNodeNotFoundError()
+        except ServicePathConflictError:
+            raise AppAssetPathConflictError()
+
+
+@console_ns.route("/apps/<string:app_id>/assets/files/<string:node_id>")
+class AppAssetFileDetailResource(Resource):
+    @setup_required
+    @login_required
+    @account_initialization_required
+    @get_app_model(mode=[AppMode.ADVANCED_CHAT, AppMode.WORKFLOW])
+    def get(self, app_model: App, node_id: str):
+        current_user, _ = current_account_with_tenant()
+        try:
+            content = AppAssetService.get_file_content(app_model, current_user.id, node_id)
+            return {"content": content.decode("utf-8", errors="replace")}
+        except ServiceNodeNotFoundError:
+            raise AppAssetNodeNotFoundError()
+
+    @console_ns.expect(console_ns.models[UpdateFileContentPayload.__name__])
+    @setup_required
+    @login_required
+    @account_initialization_required
+    @get_app_model(mode=[AppMode.ADVANCED_CHAT, AppMode.WORKFLOW])
+    def put(self, app_model: App, node_id: str):
+        current_user, _ = current_account_with_tenant()
+
+        file = request.files.get("file")
+        if file:
+            content = file.read()
+        else:
+            payload = UpdateFileContentPayload.model_validate(console_ns.payload or {})
+            content = payload.content.encode("utf-8")
+
+        try:
+            node = AppAssetService.update_file_content(app_model, current_user.id, node_id, content)
+            return node.model_dump()
+        except ServiceNodeNotFoundError:
+            raise AppAssetNodeNotFoundError()
+
+
+@console_ns.route("/apps/<string:app_id>/assets/nodes/<string:node_id>")
+class AppAssetNodeResource(Resource):
+    @setup_required
+    @login_required
+    @account_initialization_required
+    @get_app_model(mode=[AppMode.ADVANCED_CHAT, AppMode.WORKFLOW])
+    def delete(self, app_model: App, node_id: str):
+        current_user, _ = current_account_with_tenant()
+        try:
+            AppAssetService.delete_node(app_model, current_user.id, node_id)
+            return {"result": "success"}, 200
+        except ServiceNodeNotFoundError:
+            raise AppAssetNodeNotFoundError()
+
+
+@console_ns.route("/apps/<string:app_id>/assets/nodes/<string:node_id>/rename")
+class AppAssetNodeRenameResource(Resource):
+    @console_ns.expect(console_ns.models[RenameNodePayload.__name__])
+    @setup_required
+    @login_required
+    @account_initialization_required
+    @get_app_model(mode=[AppMode.ADVANCED_CHAT, AppMode.WORKFLOW])
+    def post(self, app_model: App, node_id: str):
+        current_user, _ = current_account_with_tenant()
+        payload = RenameNodePayload.model_validate(console_ns.payload or {})
+
+        try:
+            node = AppAssetService.rename_node(app_model, current_user.id, node_id, payload.name)
+            return node.model_dump()
+        except ServiceNodeNotFoundError:
+            raise AppAssetNodeNotFoundError()
+        except ServicePathConflictError:
+            raise AppAssetPathConflictError()
+
+
+@console_ns.route("/apps/<string:app_id>/assets/nodes/<string:node_id>/move")
+class AppAssetNodeMoveResource(Resource):
+    @console_ns.expect(console_ns.models[MoveNodePayload.__name__])
+    @setup_required
+    @login_required
+    @account_initialization_required
+    @get_app_model(mode=[AppMode.ADVANCED_CHAT, AppMode.WORKFLOW])
+    def post(self, app_model: App, node_id: str):
+        current_user, _ = current_account_with_tenant()
+        payload = MoveNodePayload.model_validate(console_ns.payload or {})
+
+        try:
+            node = AppAssetService.move_node(app_model, current_user.id, node_id, payload.parent_id)
+            return node.model_dump()
+        except ServiceNodeNotFoundError:
+            raise AppAssetNodeNotFoundError()
+        except AppAssetParentNotFoundError:
+            raise AppAssetNodeNotFoundError()
+        except ServicePathConflictError:
+            raise AppAssetPathConflictError()
+
+
+@console_ns.route("/apps/<string:app_id>/assets/nodes/<string:node_id>/reorder")
+class AppAssetNodeReorderResource(Resource):
+    @console_ns.expect(console_ns.models[ReorderNodePayload.__name__])
+    @setup_required
+    @login_required
+    @account_initialization_required
+    @get_app_model(mode=[AppMode.ADVANCED_CHAT, AppMode.WORKFLOW])
+    def post(self, app_model: App, node_id: str):
+        current_user, _ = current_account_with_tenant()
+        payload = ReorderNodePayload.model_validate(console_ns.payload or {})
+
+        try:
+            node = AppAssetService.reorder_node(app_model, current_user.id, node_id, payload.after_node_id)
+            return node.model_dump()
+        except ServiceNodeNotFoundError:
+            raise AppAssetNodeNotFoundError()
+
+
+@console_ns.route("/apps/<string:app_id>/assets/publish")
+class AppAssetPublishResource(Resource):
+    @setup_required
+    @login_required
+    @account_initialization_required
+    @get_app_model(mode=[AppMode.ADVANCED_CHAT, AppMode.WORKFLOW])
+    def post(self, app_model: App):
+        current_user, _ = current_account_with_tenant()
+        published = AppAssetService.publish(app_model, current_user.id)
+        return {
+            "id": published.id,
+            "version": published.version,
+            "asset_tree": published.asset_tree.model_dump(),
+        }, 201
+
+
+@console_ns.route("/apps/<string:app_id>/assets/files/<string:node_id>/download-url")
+class AppAssetFileDownloadUrlResource(Resource):
+    @setup_required
+    @login_required
+    @account_initialization_required
+    @get_app_model(mode=[AppMode.ADVANCED_CHAT, AppMode.WORKFLOW])
+    def get(self, app_model: App, node_id: str):
+        current_user, _ = current_account_with_tenant()
+        try:
+            download_url = AppAssetService.get_file_download_url(app_model, current_user.id, node_id)
+            return {"download_url": download_url}
+        except ServiceNodeNotFoundError:
+            raise AppAssetNodeNotFoundError()

api/controllers/console/app/error.py

@@ -110,8 +110,6 @@ class TracingConfigCheckError(BaseHTTPException):
 
 
 class InvokeRateLimitError(BaseHTTPException):
-    """Raised when the Invoke returns rate limit error."""
-
     error_code = "rate_limit_error"
     description = "Rate Limit Error"
     code = 429
@@ -121,3 +119,21 @@ class NeedAddIdsError(BaseHTTPException):
     error_code = "need_add_ids"
     description = "Need to add ids."
     code = 400
+
+
+class AppAssetNodeNotFoundError(BaseHTTPException):
+    error_code = "app_asset_node_not_found"
+    description = "App asset node not found."
+    code = 404
+
+
+class AppAssetFileRequiredError(BaseHTTPException):
+    error_code = "app_asset_file_required"
+    description = "File is required."
+    code = 400
+
+
+class AppAssetPathConflictError(BaseHTTPException):
+    error_code = "app_asset_path_conflict"
+    description = "Path already exists."
+    code = 409

api/controllers/console/app/generator.py

@@ -55,6 +55,35 @@ class InstructionTemplatePayload(BaseModel):
     type: str = Field(..., description="Instruction template type")
 
 
+class ContextGeneratePayload(BaseModel):
+    """Payload for generating extractor code node."""
+
+    workflow_id: str = Field(..., description="Workflow ID")
+    node_id: str = Field(..., description="Current tool/llm node ID")
+    parameter_name: str = Field(..., description="Parameter name to generate code for")
+    language: str = Field(default="python3", description="Code language (python3/javascript)")
+    prompt_messages: list[dict[str, Any]] = Field(
+        ..., description="Multi-turn conversation history, last message is the current instruction"
+    )
+    model_config_data: dict[str, Any] = Field(..., alias="model_config", description="Model configuration")
+
+
+class SuggestedQuestionsPayload(BaseModel):
+    """Payload for generating suggested questions."""
+
+    workflow_id: str = Field(..., description="Workflow ID")
+    node_id: str = Field(..., description="Current tool/llm node ID")
+    parameter_name: str = Field(..., description="Parameter name")
+    language: str = Field(
+        default="English", description="Language for generated questions (e.g. English, Chinese, Japanese)"
+    )
+    model_config_data: dict[str, Any] | None = Field(
+        default=None,
+        alias="model_config",
+        description="Model configuration (optional, uses system default if not provided)",
+    )
+
+
 def reg(cls: type[BaseModel]):
     console_ns.schema_model(cls.__name__, cls.model_json_schema(ref_template=DEFAULT_REF_TEMPLATE_SWAGGER_2_0))
 
@@ -64,6 +93,8 @@ reg(RuleCodeGeneratePayload)
 reg(RuleStructuredOutputPayload)
 reg(InstructionGeneratePayload)
 reg(InstructionTemplatePayload)
+reg(ContextGeneratePayload)
+reg(SuggestedQuestionsPayload)
 
 
 @console_ns.route("/rule-generate")
@@ -278,3 +309,74 @@ class InstructionGenerationTemplateApi(Resource):
                 return {"data": INSTRUCTION_GENERATE_TEMPLATE_CODE}
             case _:
                 raise ValueError(f"Invalid type: {args.type}")
+
+
+@console_ns.route("/context-generate")
+class ContextGenerateApi(Resource):
+    @console_ns.doc("generate_with_context")
+    @console_ns.doc(description="Generate with multi-turn conversation context")
+    @console_ns.expect(console_ns.models[ContextGeneratePayload.__name__])
+    @console_ns.response(200, "Content generated successfully")
+    @console_ns.response(400, "Invalid request parameters or workflow not found")
+    @console_ns.response(402, "Provider quota exceeded")
+    @setup_required
+    @login_required
+    @account_initialization_required
+    def post(self):
+        from core.llm_generator.utils import deserialize_prompt_messages
+
+        args = ContextGeneratePayload.model_validate(console_ns.payload)
+        _, current_tenant_id = current_account_with_tenant()
+
+        prompt_messages = deserialize_prompt_messages(args.prompt_messages)
+
+        try:
+            return LLMGenerator.generate_with_context(
+                tenant_id=current_tenant_id,
+                workflow_id=args.workflow_id,
+                node_id=args.node_id,
+                parameter_name=args.parameter_name,
+                language=args.language,
+                prompt_messages=prompt_messages,
+                model_config=args.model_config_data,
+            )
+        except ProviderTokenNotInitError as ex:
+            raise ProviderNotInitializeError(ex.description)
+        except QuotaExceededError:
+            raise ProviderQuotaExceededError()
+        except ModelCurrentlyNotSupportError:
+            raise ProviderModelCurrentlyNotSupportError()
+        except InvokeError as e:
+            raise CompletionRequestError(e.description)
+
+
+@console_ns.route("/context-generate/suggested-questions")
+class SuggestedQuestionsApi(Resource):
+    @console_ns.doc("generate_suggested_questions")
+    @console_ns.doc(description="Generate suggested questions for context generation")
+    @console_ns.expect(console_ns.models[SuggestedQuestionsPayload.__name__])
+    @console_ns.response(200, "Questions generated successfully")
+    @setup_required
+    @login_required
+    @account_initialization_required
+    def post(self):
+        args = SuggestedQuestionsPayload.model_validate(console_ns.payload)
+        _, current_tenant_id = current_account_with_tenant()
+
+        try:
+            return LLMGenerator.generate_suggested_questions(
+                tenant_id=current_tenant_id,
+                workflow_id=args.workflow_id,
+                node_id=args.node_id,
+                parameter_name=args.parameter_name,
+                language=args.language,
+                model_config=args.model_config_data,
+            )
+        except ProviderTokenNotInitError as ex:
+            raise ProviderNotInitializeError(ex.description)
+        except QuotaExceededError:
+            raise ProviderQuotaExceededError()
+        except ModelCurrentlyNotSupportError:
+            raise ProviderModelCurrentlyNotSupportError()
+        except InvokeError as e:
+            raise CompletionRequestError(e.description)

api/controllers/console/app/message.py

@@ -202,6 +202,7 @@ message_detail_model = console_ns.model(
         "status": fields.String,
         "error": fields.String,
         "parent_message_id": fields.String,
+        "generation_detail": fields.Raw,
     },
 )
 

api/controllers/console/app/workflow.py

@@ -46,6 +46,8 @@ from models.workflow import Workflow
 from services.app_generate_service import AppGenerateService
 from services.errors.app import WorkflowHashNotEqualError
 from services.errors.llm import InvokeRateLimitError
+from services.workflow.entities import NestedNodeGraphRequest, NestedNodeParameterSchema
+from services.workflow.nested_node_graph_service import NestedNodeGraphService
 from services.workflow_service import DraftWorkflowDeletionError, WorkflowInUseError, WorkflowService
 
 logger = logging.getLogger(__name__)
@@ -188,6 +190,15 @@ class DraftWorkflowTriggerRunAllPayload(BaseModel):
     node_ids: list[str]
 
 
+class NestedNodeGraphPayload(BaseModel):
+    """Request payload for generating nested node graph."""
+
+    parent_node_id: str = Field(description="ID of the parent node that uses the extracted value")
+    parameter_key: str = Field(description="Key of the parameter being extracted")
+    context_source: list[str] = Field(description="Variable selector for the context source")
+    parameter_schema: dict[str, Any] = Field(description="Schema of the parameter to extract")
+
+
 def reg(cls: type[BaseModel]):
     console_ns.schema_model(cls.__name__, cls.model_json_schema(ref_template=DEFAULT_REF_TEMPLATE_SWAGGER_2_0))
 
@@ -205,6 +216,7 @@ reg(WorkflowListQuery)
 reg(WorkflowUpdatePayload)
 reg(DraftWorkflowTriggerRunPayload)
 reg(DraftWorkflowTriggerRunAllPayload)
+reg(NestedNodeGraphPayload)
 
 
 # TODO(QuantumGhost): Refactor existing node run API to handle file parameter parsing
@@ -1166,3 +1178,54 @@ class DraftWorkflowTriggerRunAllApi(Resource):
                     "status": "error",
                 }
             ), 400
+
+
+@console_ns.route("/apps/<uuid:app_id>/workflows/draft/nested-node-graph")
+class NestedNodeGraphApi(Resource):
+    """
+    API for generating Nested Node LLM graph structures.
+
+    This endpoint creates a complete graph structure containing an LLM node
+    configured to extract values from list[PromptMessage] variables.
+    """
+
+    @console_ns.doc("generate_nested_node_graph")
+    @console_ns.doc(description="Generate a Nested Node LLM graph structure")
+    @console_ns.doc(params={"app_id": "Application ID"})
+    @console_ns.expect(console_ns.models[NestedNodeGraphPayload.__name__])
+    @console_ns.response(200, "Nested node graph generated successfully")
+    @console_ns.response(400, "Invalid request parameters")
+    @console_ns.response(403, "Permission denied")
+    @setup_required
+    @login_required
+    @account_initialization_required
+    @get_app_model(mode=[AppMode.ADVANCED_CHAT, AppMode.WORKFLOW])
+    @edit_permission_required
+    def post(self, app_model: App):
+        """
+        Generate a Nested Node LLM graph structure.
+
+        Returns a complete graph structure containing a single LLM node
+        configured for extracting values from list[PromptMessage] context.
+        """
+
+        payload = NestedNodeGraphPayload.model_validate(console_ns.payload or {})
+
+        parameter_schema = NestedNodeParameterSchema(
+            name=payload.parameter_schema.get("name", payload.parameter_key),
+            type=payload.parameter_schema.get("type", "string"),
+            description=payload.parameter_schema.get("description", ""),
+        )
+
+        request = NestedNodeGraphRequest(
+            parent_node_id=payload.parent_node_id,
+            parameter_key=payload.parameter_key,
+            context_source=payload.context_source,
+            parameter_schema=parameter_schema,
+        )
+
+        with Session(db.engine) as session:
+            service = NestedNodeGraphService(session)
+            response = service.generate_nested_node_graph(tenant_id=app_model.tenant_id, request=request)
+
+        return response.model_dump()

api/controllers/console/app/workflow_draft_variable.py

@@ -16,14 +16,15 @@ from controllers.console.app.wraps import get_app_model
 from controllers.console.wraps import account_initialization_required, edit_permission_required, setup_required
 from controllers.web.error import InvalidArgumentError, NotFoundError
 from core.file import helpers as file_helpers
+from core.sandbox.manager import SandboxManager
 from core.variables.segment_group import SegmentGroup
-from core.variables.segments import ArrayFileSegment, FileSegment, Segment
+from core.variables.segments import ArrayFileSegment, ArrayPromptMessageSegment, FileSegment, Segment
 from core.variables.types import SegmentType
 from core.workflow.constants import CONVERSATION_VARIABLE_NODE_ID, SYSTEM_VARIABLE_NODE_ID
 from extensions.ext_database import db
 from factories.file_factory import build_from_mapping, build_from_mappings
 from factories.variable_factory import build_segment_with_type
-from libs.login import login_required
+from libs.login import current_account_with_tenant, login_required
 from models import App, AppMode
 from models.workflow import WorkflowDraftVariable
 from services.workflow_draft_variable_service import WorkflowDraftVariableList, WorkflowDraftVariableService
@@ -58,6 +59,8 @@ def _convert_values_to_json_serializable_object(value: Segment):
         return value.value.model_dump()
     elif isinstance(value, ArrayFileSegment):
         return [i.model_dump() for i in value.value]
+    elif isinstance(value, ArrayPromptMessageSegment):
+        return value.to_object()
     elif isinstance(value, SegmentGroup):
         return [_convert_values_to_json_serializable_object(i) for i in value.value]
     else:
@@ -247,6 +250,9 @@ class WorkflowVariableCollectionApi(Resource):
     @console_ns.response(204, "Workflow variables deleted successfully")
     @_api_prerequisite
     def delete(self, app_model: App):
+        # FIXME(Mairuis): move to SandboxArtifactService
+        current_user, _ = current_account_with_tenant()
+        SandboxManager.delete_storage(app_model.tenant_id, current_user.id)
         draft_var_srv = WorkflowDraftVariableService(
             session=db.session(),
         )

api/controllers/console/workspace/dsl.py

@@ -0,0 +1,65 @@
+import json
+
+import httpx
+import yaml
+from flask_restx import Resource, reqparse
+from sqlalchemy.orm import Session
+from werkzeug.exceptions import Forbidden
+
+from controllers.console import console_ns
+from controllers.console.wraps import account_initialization_required, setup_required
+from core.plugin.impl.exc import PluginPermissionDeniedError
+from extensions.ext_database import db
+from libs.login import current_account_with_tenant, login_required
+from models.model import App
+from models.workflow import Workflow
+from services.app_dsl_service import AppDslService
+
+
+@console_ns.route("/workspaces/current/dsl/predict")
+class DSLPredictApi(Resource):
+    @setup_required
+    @login_required
+    @account_initialization_required
+    def post(self):
+        user, _ = current_account_with_tenant()
+        if not user.is_admin_or_owner:
+            raise Forbidden()
+
+        parser = (
+            reqparse.RequestParser()
+            .add_argument("app_id", type=str, required=True, location="json")
+            .add_argument("current_node_id", type=str, required=True, location="json")
+        )
+        args = parser.parse_args()
+
+        app_id: str = args["app_id"]
+        current_node_id: str = args["current_node_id"]
+
+        with Session(db.engine) as session:
+            app = session.query(App).filter_by(id=app_id).first()
+            workflow = session.query(Workflow).filter_by(app_id=app_id, version=Workflow.VERSION_DRAFT).first()
+
+        if not app:
+            raise ValueError("App not found")
+        if not workflow:
+            raise ValueError("Workflow not found")
+
+        try:
+            i = 0
+            for node_id, _ in workflow.walk_nodes():
+                if node_id == current_node_id:
+                    break
+                i += 1
+
+            dsl = yaml.safe_load(AppDslService.export_dsl(app_model=app))
+
+            response = httpx.post(
+                "http://spark-832c:8000/predict",
+                json={"graph_data": dsl, "source_node_index": i},
+            )
+            return {
+                "nodes": json.loads(response.json()),
+            }
+        except PluginPermissionDeniedError as e:
+            raise ValueError(e.description) from e

api/controllers/console/workspace/sandbox_providers.py

@@ -0,0 +1,103 @@
+import logging
+
+from flask_restx import Resource, fields, reqparse
+
+from controllers.console import console_ns
+from controllers.console.wraps import account_initialization_required, setup_required
+from core.model_runtime.utils.encoders import jsonable_encoder
+from libs.login import current_account_with_tenant, login_required
+from services.sandbox.sandbox_provider_service import SandboxProviderService
+
+logger = logging.getLogger(__name__)
+
+
+@console_ns.route("/workspaces/current/sandbox-providers")
+class SandboxProviderListApi(Resource):
+    @console_ns.doc("list_sandbox_providers")
+    @console_ns.doc(description="Get list of available sandbox providers with configuration status")
+    @console_ns.response(200, "Success", fields.List(fields.Raw(description="Sandbox provider information")))
+    @setup_required
+    @login_required
+    @account_initialization_required
+    def get(self):
+        _, current_tenant_id = current_account_with_tenant()
+        providers = SandboxProviderService.list_providers(current_tenant_id)
+        return jsonable_encoder([p.model_dump() for p in providers])
+
+
+config_parser = reqparse.RequestParser()
+config_parser.add_argument("config", type=dict, required=True, location="json")
+config_parser.add_argument("activate", type=bool, required=False, default=False, location="json")
+
+
+@console_ns.route("/workspaces/current/sandbox-provider/<string:provider_type>/config")
+class SandboxProviderConfigApi(Resource):
+    @console_ns.doc("save_sandbox_provider_config")
+    @console_ns.doc(description="Save or update configuration for a sandbox provider")
+    @console_ns.expect(config_parser)
+    @console_ns.response(200, "Success")
+    @setup_required
+    @login_required
+    @account_initialization_required
+    def post(self, provider_type: str):
+        _, current_tenant_id = current_account_with_tenant()
+        args = config_parser.parse_args()
+
+        try:
+            result = SandboxProviderService.save_config(
+                tenant_id=current_tenant_id,
+                provider_type=provider_type,
+                config=args["config"],
+                activate=args["activate"],
+            )
+            return result
+        except ValueError as e:
+            return {"message": str(e)}, 400
+
+    @console_ns.doc("delete_sandbox_provider_config")
+    @console_ns.doc(description="Delete configuration for a sandbox provider")
+    @console_ns.response(200, "Success")
+    @setup_required
+    @login_required
+    @account_initialization_required
+    def delete(self, provider_type: str):
+        _, current_tenant_id = current_account_with_tenant()
+
+        try:
+            result = SandboxProviderService.delete_config(
+                tenant_id=current_tenant_id,
+                provider_type=provider_type,
+            )
+            return result
+        except ValueError as e:
+            return {"message": str(e)}, 400
+
+
+activate_parser = reqparse.RequestParser()
+activate_parser.add_argument("type", type=str, required=True, location="json")
+
+
+@console_ns.route("/workspaces/current/sandbox-provider/<string:provider_type>/activate")
+class SandboxProviderActivateApi(Resource):
+    """Activate a sandbox provider."""
+
+    @console_ns.doc("activate_sandbox_provider")
+    @console_ns.doc(description="Activate a sandbox provider for the current workspace")
+    @console_ns.response(200, "Success")
+    @setup_required
+    @login_required
+    @account_initialization_required
+    def post(self, provider_type: str):
+        """Activate a sandbox provider."""
+        _, current_tenant_id = current_account_with_tenant()
+
+        try:
+            args = activate_parser.parse_args()
+            result = SandboxProviderService.activate_provider(
+                tenant_id=current_tenant_id,
+                provider_type=provider_type,
+                type=args["type"],
+            )
+            return result
+        except ValueError as e:
+            return {"message": str(e)}, 400

api/controllers/files/__init__.py

@@ -14,7 +14,7 @@ api = ExternalApi(
 
 files_ns = Namespace("files", description="File operations", path="/")
 
-from . import image_preview, tool_files, upload
+from . import image_preview, storage_download, tool_files, upload
 
 api.add_namespace(files_ns)
 
@@ -23,6 +23,7 @@ __all__ = [
     "bp",
     "files_ns",
     "image_preview",
+    "storage_download",
     "tool_files",
     "upload",
 ]

api/controllers/files/storage_download.py

@@ -0,0 +1,56 @@
+from urllib.parse import quote, unquote
+
+from flask import Response, request
+from flask_restx import Resource
+from pydantic import BaseModel, Field
+from werkzeug.exceptions import Forbidden, NotFound
+
+from controllers.files import files_ns
+from extensions.ext_storage import storage
+from extensions.storage.file_presign_storage import FilePresignStorage
+
+DEFAULT_REF_TEMPLATE_SWAGGER_2_0 = "#/definitions/{model}"
+
+
+class StorageDownloadQuery(BaseModel):
+    timestamp: str = Field(..., description="Unix timestamp used in the signature")
+    nonce: str = Field(..., description="Random string for signature")
+    sign: str = Field(..., description="HMAC signature")
+
+
+files_ns.schema_model(
+    StorageDownloadQuery.__name__,
+    StorageDownloadQuery.model_json_schema(ref_template=DEFAULT_REF_TEMPLATE_SWAGGER_2_0),
+)
+
+
+@files_ns.route("/storage/<path:filename>/download")
+class StorageFileDownloadApi(Resource):
+    def get(self, filename: str):
+        filename = unquote(filename)
+
+        args = StorageDownloadQuery.model_validate(request.args.to_dict(flat=True))
+
+        if not FilePresignStorage.verify_signature(
+            filename=filename,
+            timestamp=args.timestamp,
+            nonce=args.nonce,
+            sign=args.sign,
+        ):
+            raise Forbidden("Invalid or expired download link")
+
+        try:
+            generator = storage.load_stream(filename)
+        except FileNotFoundError:
+            raise NotFound("File not found")
+
+        encoded_filename = quote(filename.split("/")[-1])
+
+        return Response(
+            generator,
+            mimetype="application/octet-stream",
+            direct_passthrough=True,
+            headers={
+                "Content-Disposition": f"attachment; filename*=UTF-8''{encoded_filename}",
+            },
+        )

api/controllers/inner_api/plugin/plugin.py

@@ -448,3 +448,53 @@ class PluginFetchAppInfoApi(Resource):
         return BaseBackwardsInvocationResponse(
             data=PluginAppBackwardsInvocation.fetch_app_info(payload.app_id, tenant_model.id)
         ).model_dump()
+
+
+@inner_api_ns.route("/fetch/tools/list")
+class PluginFetchToolsListApi(Resource):
+    @get_user_tenant
+    @setup_required
+    @plugin_inner_api_only
+    @inner_api_ns.doc("plugin_fetch_tools_list")
+    @inner_api_ns.doc(description="Fetch all available tools through plugin interface")
+    @inner_api_ns.doc(
+        responses={
+            200: "Tools list retrieved successfully",
+            401: "Unauthorized - invalid API key",
+            404: "Service not available",
+        }
+    )
+    def post(self, user_model: Account | EndUser, tenant_model: Tenant):
+        from sqlalchemy.orm import Session
+
+        from extensions.ext_database import db
+        from services.tools.api_tools_manage_service import ApiToolManageService
+        from services.tools.builtin_tools_manage_service import BuiltinToolManageService
+        from services.tools.mcp_tools_manage_service import MCPToolManageService
+        from services.tools.workflow_tools_manage_service import WorkflowToolManageService
+
+        providers = []
+
+        # Get builtin tools
+        builtin_providers = BuiltinToolManageService.list_builtin_tools(user_model.id, tenant_model.id)
+        for provider in builtin_providers:
+            providers.append(provider.to_dict())
+
+        # Get API tools
+        api_providers = ApiToolManageService.list_api_tools(tenant_model.id)
+        for provider in api_providers:
+            providers.append(provider.to_dict())
+
+        # Get workflow tools
+        workflow_providers = WorkflowToolManageService.list_tenant_workflow_tools(user_model.id, tenant_model.id)
+        for provider in workflow_providers:
+            providers.append(provider.to_dict())
+
+        # Get MCP tools
+        with Session(db.engine) as session:
+            mcp_service = MCPToolManageService(session)
+            mcp_providers = mcp_service.list_providers(tenant_id=tenant_model.id, for_list=True)
+            for provider in mcp_providers:
+                providers.append(provider.to_dict())
+
+        return BaseBackwardsInvocationResponse(data={"providers": providers}).model_dump()

api/controllers/inner_api/plugin/wraps.py

@@ -75,7 +75,6 @@ def get_user_tenant(view_func: Callable[P, R]):
     @wraps(view_func)
     def decorated_view(*args: P.args, **kwargs: P.kwargs):
         payload = TenantUserPayload.model_validate(request.get_json(silent=True) or {})
-
         user_id = payload.user_id
         tenant_id = payload.tenant_id
 

api/controllers/inner_api/wraps.py

@@ -5,14 +5,15 @@ from hashlib import sha1
 from hmac import new as hmac_new
 from typing import ParamSpec, TypeVar
 
-P = ParamSpec("P")
-R = TypeVar("R")
 from flask import abort, request
 
 from configs import dify_config
 from extensions.ext_database import db
 from models.model import EndUser
 
+P = ParamSpec("P")
+R = TypeVar("R")
+
 
 def billing_inner_api_only(view: Callable[P, R]):
     @wraps(view)
@@ -88,11 +89,11 @@ def plugin_inner_api_only(view: Callable[P, R]):
         if not dify_config.PLUGIN_DAEMON_KEY:
             abort(404)
 
-        # get header 'X-Inner-Api-Key'
+        # validate using inner api key
         inner_api_key = request.headers.get("X-Inner-Api-Key")
-        if not inner_api_key or inner_api_key != dify_config.INNER_API_KEY_FOR_PLUGIN:
-            abort(404)
+        if inner_api_key and inner_api_key == dify_config.INNER_API_KEY_FOR_PLUGIN:
+            return view(*args, **kwargs)
 
-        return view(*args, **kwargs)
+        abort(401)
 
     return decorated

api/core/agent/agent_app_runner.py

@@ -0,0 +1,380 @@
+import logging
+from collections.abc import Generator
+from copy import deepcopy
+from typing import Any
+
+from core.agent.base_agent_runner import BaseAgentRunner
+from core.agent.entities import AgentEntity, AgentLog, AgentResult
+from core.agent.patterns.strategy_factory import StrategyFactory
+from core.app.apps.base_app_queue_manager import PublishFrom
+from core.app.entities.queue_entities import QueueAgentThoughtEvent, QueueMessageEndEvent, QueueMessageFileEvent
+from core.file import file_manager
+from core.model_runtime.entities import (
+    AssistantPromptMessage,
+    LLMResult,
+    LLMResultChunk,
+    LLMUsage,
+    PromptMessage,
+    PromptMessageContentType,
+    SystemPromptMessage,
+    TextPromptMessageContent,
+    UserPromptMessage,
+)
+from core.model_runtime.entities.message_entities import ImagePromptMessageContent, PromptMessageContentUnionTypes
+from core.prompt.agent_history_prompt_transform import AgentHistoryPromptTransform
+from core.tools.__base.tool import Tool
+from core.tools.entities.tool_entities import ToolInvokeMeta
+from core.tools.tool_engine import ToolEngine
+from models.model import Message
+
+logger = logging.getLogger(__name__)
+
+
+class AgentAppRunner(BaseAgentRunner):
+    def _create_tool_invoke_hook(self, message: Message):
+        """
+        Create a tool invoke hook that uses ToolEngine.agent_invoke.
+        This hook handles file creation and returns proper meta information.
+        """
+        # Get trace manager from app generate entity
+        trace_manager = self.application_generate_entity.trace_manager
+
+        def tool_invoke_hook(
+            tool: Tool, tool_args: dict[str, Any], tool_name: str
+        ) -> tuple[str, list[str], ToolInvokeMeta]:
+            """Hook that uses agent_invoke for proper file and meta handling."""
+            tool_invoke_response, message_files, tool_invoke_meta = ToolEngine.agent_invoke(
+                tool=tool,
+                tool_parameters=tool_args,
+                user_id=self.user_id,
+                tenant_id=self.tenant_id,
+                message=message,
+                invoke_from=self.application_generate_entity.invoke_from,
+                agent_tool_callback=self.agent_callback,
+                trace_manager=trace_manager,
+                app_id=self.application_generate_entity.app_config.app_id,
+                message_id=message.id,
+                conversation_id=self.conversation.id,
+            )
+
+            # Publish files and track IDs
+            for message_file_id in message_files:
+                self.queue_manager.publish(
+                    QueueMessageFileEvent(message_file_id=message_file_id),
+                    PublishFrom.APPLICATION_MANAGER,
+                )
+                self._current_message_file_ids.append(message_file_id)
+
+            return tool_invoke_response, message_files, tool_invoke_meta
+
+        return tool_invoke_hook
+
+    def run(self, message: Message, query: str, **kwargs: Any) -> Generator[LLMResultChunk, None, None]:
+        """
+        Run Agent application
+        """
+        self.query = query
+        app_generate_entity = self.application_generate_entity
+
+        app_config = self.app_config
+        assert app_config is not None, "app_config is required"
+        assert app_config.agent is not None, "app_config.agent is required"
+
+        # convert tools into ModelRuntime Tool format
+        tool_instances, _ = self._init_prompt_tools()
+
+        assert app_config.agent
+
+        # Create tool invoke hook for agent_invoke
+        tool_invoke_hook = self._create_tool_invoke_hook(message)
+
+        # Get instruction for ReAct strategy
+        instruction = self.app_config.prompt_template.simple_prompt_template or ""
+
+        # Use factory to create appropriate strategy
+        strategy = StrategyFactory.create_strategy(
+            model_features=self.model_features,
+            model_instance=self.model_instance,
+            tools=list(tool_instances.values()),
+            files=list(self.files),
+            max_iterations=app_config.agent.max_iteration,
+            context=self.build_execution_context(),
+            agent_strategy=self.config.strategy,
+            tool_invoke_hook=tool_invoke_hook,
+            instruction=instruction,
+        )
+
+        # Initialize state variables
+        current_agent_thought_id = None
+        has_published_thought = False
+        current_tool_name: str | None = None
+        self._current_message_file_ids: list[str] = []
+
+        # organize prompt messages
+        prompt_messages = self._organize_prompt_messages()
+
+        # Run strategy
+        generator = strategy.run(
+            prompt_messages=prompt_messages,
+            model_parameters=app_generate_entity.model_conf.parameters,
+            stop=app_generate_entity.model_conf.stop,
+            stream=True,
+        )
+
+        # Consume generator and collect result
+        result: AgentResult | None = None
+        try:
+            while True:
+                try:
+                    output = next(generator)
+                except StopIteration as e:
+                    # Generator finished, get the return value
+                    result = e.value
+                    break
+
+                if isinstance(output, LLMResultChunk):
+                    # Handle LLM chunk
+                    if current_agent_thought_id and not has_published_thought:
+                        self.queue_manager.publish(
+                            QueueAgentThoughtEvent(agent_thought_id=current_agent_thought_id),
+                            PublishFrom.APPLICATION_MANAGER,
+                        )
+                        has_published_thought = True
+
+                    yield output
+
+                elif isinstance(output, AgentLog):
+                    # Handle Agent Log using log_type for type-safe dispatch
+                    if output.status == AgentLog.LogStatus.START:
+                        if output.log_type == AgentLog.LogType.ROUND:
+                            # Start of a new round
+                            message_file_ids: list[str] = []
+                            current_agent_thought_id = self.create_agent_thought(
+                                message_id=message.id,
+                                message="",
+                                tool_name="",
+                                tool_input="",
+                                messages_ids=message_file_ids,
+                            )
+                            has_published_thought = False
+
+                        elif output.log_type == AgentLog.LogType.TOOL_CALL:
+                            if current_agent_thought_id is None:
+                                continue
+
+                            # Tool call start - extract data from structured fields
+                            current_tool_name = output.data.get("tool_name", "")
+                            tool_input = output.data.get("tool_args", {})
+
+                            self.save_agent_thought(
+                                agent_thought_id=current_agent_thought_id,
+                                tool_name=current_tool_name,
+                                tool_input=tool_input,
+                                thought=None,
+                                observation=None,
+                                tool_invoke_meta=None,
+                                answer=None,
+                                messages_ids=[],
+                            )
+                            self.queue_manager.publish(
+                                QueueAgentThoughtEvent(agent_thought_id=current_agent_thought_id),
+                                PublishFrom.APPLICATION_MANAGER,
+                            )
+
+                    elif output.status == AgentLog.LogStatus.SUCCESS:
+                        if output.log_type == AgentLog.LogType.THOUGHT:
+                            if current_agent_thought_id is None:
+                                continue
+
+                            thought_text = output.data.get("thought")
+                            self.save_agent_thought(
+                                agent_thought_id=current_agent_thought_id,
+                                tool_name=None,
+                                tool_input=None,
+                                thought=thought_text,
+                                observation=None,
+                                tool_invoke_meta=None,
+                                answer=None,
+                                messages_ids=[],
+                            )
+                            self.queue_manager.publish(
+                                QueueAgentThoughtEvent(agent_thought_id=current_agent_thought_id),
+                                PublishFrom.APPLICATION_MANAGER,
+                            )
+
+                        elif output.log_type == AgentLog.LogType.TOOL_CALL:
+                            if current_agent_thought_id is None:
+                                continue
+
+                            # Tool call finished
+                            tool_output = output.data.get("output")
+                            # Get meta from strategy output (now properly populated)
+                            tool_meta = output.data.get("meta")
+
+                            # Wrap tool_meta with tool_name as key (required by agent_service)
+                            if tool_meta and current_tool_name:
+                                tool_meta = {current_tool_name: tool_meta}
+
+                            self.save_agent_thought(
+                                agent_thought_id=current_agent_thought_id,
+                                tool_name=None,
+                                tool_input=None,
+                                thought=None,
+                                observation=tool_output,
+                                tool_invoke_meta=tool_meta,
+                                answer=None,
+                                messages_ids=self._current_message_file_ids,
+                            )
+                            # Clear message file ids after saving
+                            self._current_message_file_ids = []
+                            current_tool_name = None
+
+                            self.queue_manager.publish(
+                                QueueAgentThoughtEvent(agent_thought_id=current_agent_thought_id),
+                                PublishFrom.APPLICATION_MANAGER,
+                            )
+
+                        elif output.log_type == AgentLog.LogType.ROUND:
+                            if current_agent_thought_id is None:
+                                continue
+
+                            # Round finished - save LLM usage and answer
+                            llm_usage = output.metadata.get(AgentLog.LogMetadata.LLM_USAGE)
+                            llm_result = output.data.get("llm_result")
+                            final_answer = output.data.get("final_answer")
+
+                            self.save_agent_thought(
+                                agent_thought_id=current_agent_thought_id,
+                                tool_name=None,
+                                tool_input=None,
+                                thought=llm_result,
+                                observation=None,
+                                tool_invoke_meta=None,
+                                answer=final_answer,
+                                messages_ids=[],
+                                llm_usage=llm_usage,
+                            )
+                            self.queue_manager.publish(
+                                QueueAgentThoughtEvent(agent_thought_id=current_agent_thought_id),
+                                PublishFrom.APPLICATION_MANAGER,
+                            )
+
+        except Exception:
+            # Re-raise any other exceptions
+            raise
+
+        # Process final result
+        if isinstance(result, AgentResult):
+            final_answer = result.text
+            usage = result.usage or LLMUsage.empty_usage()
+
+            # Publish end event
+            self.queue_manager.publish(
+                QueueMessageEndEvent(
+                    llm_result=LLMResult(
+                        model=self.model_instance.model,
+                        prompt_messages=prompt_messages,
+                        message=AssistantPromptMessage(content=final_answer),
+                        usage=usage,
+                        system_fingerprint="",
+                    )
+                ),
+                PublishFrom.APPLICATION_MANAGER,
+            )
+
+    def _init_system_message(self, prompt_template: str, prompt_messages: list[PromptMessage]) -> list[PromptMessage]:
+        """
+        Initialize system message
+        """
+        if not prompt_template:
+            return prompt_messages or []
+
+        prompt_messages = prompt_messages or []
+
+        if prompt_messages and isinstance(prompt_messages[0], SystemPromptMessage):
+            prompt_messages[0] = SystemPromptMessage(content=prompt_template)
+            return prompt_messages
+
+        if not prompt_messages:
+            return [SystemPromptMessage(content=prompt_template)]
+
+        prompt_messages.insert(0, SystemPromptMessage(content=prompt_template))
+        return prompt_messages
+
+    def _organize_user_query(self, query: str, prompt_messages: list[PromptMessage]) -> list[PromptMessage]:
+        """
+        Organize user query
+        """
+        if self.files:
+            # get image detail config
+            image_detail_config = (
+                self.application_generate_entity.file_upload_config.image_config.detail
+                if (
+                    self.application_generate_entity.file_upload_config
+                    and self.application_generate_entity.file_upload_config.image_config
+                )
+                else None
+            )
+            image_detail_config = image_detail_config or ImagePromptMessageContent.DETAIL.LOW
+
+            prompt_message_contents: list[PromptMessageContentUnionTypes] = []
+            for file in self.files:
+                prompt_message_contents.append(
+                    file_manager.to_prompt_message_content(
+                        file,
+                        image_detail_config=image_detail_config,
+                    )
+                )
+            prompt_message_contents.append(TextPromptMessageContent(data=query))
+
+            prompt_messages.append(UserPromptMessage(content=prompt_message_contents))
+        else:
+            prompt_messages.append(UserPromptMessage(content=query))
+
+        return prompt_messages
+
+    def _clear_user_prompt_image_messages(self, prompt_messages: list[PromptMessage]) -> list[PromptMessage]:
+        """
+        As for now, gpt supports both fc and vision at the first iteration.
+        We need to remove the image messages from the prompt messages at the first iteration.
+        """
+        prompt_messages = deepcopy(prompt_messages)
+
+        for prompt_message in prompt_messages:
+            if isinstance(prompt_message, UserPromptMessage):
+                if isinstance(prompt_message.content, list):
+                    prompt_message.content = "\n".join(
+                        [
+                            content.data
+                            if content.type == PromptMessageContentType.TEXT
+                            else "[image]"
+                            if content.type == PromptMessageContentType.IMAGE
+                            else "[file]"
+                            for content in prompt_message.content
+                        ]
+                    )
+
+        return prompt_messages
+
+    def _organize_prompt_messages(self):
+        # For ReAct strategy, use the agent prompt template
+        if self.config.strategy == AgentEntity.Strategy.CHAIN_OF_THOUGHT and self.config.prompt:
+            prompt_template = self.config.prompt.first_prompt
+        else:
+            prompt_template = self.app_config.prompt_template.simple_prompt_template or ""
+
+        self.history_prompt_messages = self._init_system_message(prompt_template, self.history_prompt_messages)
+        query_prompt_messages = self._organize_user_query(self.query or "", [])
+
+        self.history_prompt_messages = AgentHistoryPromptTransform(
+            model_config=self.model_config,
+            prompt_messages=[*query_prompt_messages, *self._current_thoughts],
+            history_messages=self.history_prompt_messages,
+            memory=self.memory,
+        ).get_prompt()
+
+        prompt_messages = [*self.history_prompt_messages, *query_prompt_messages, *self._current_thoughts]
+        if len(self._current_thoughts) != 0:
+            # clear messages after the first iteration
+            prompt_messages = self._clear_user_prompt_image_messages(prompt_messages)
+        return prompt_messages

api/core/agent/base_agent_runner.py

@@ -6,7 +6,7 @@ from typing import Union, cast
 
 from sqlalchemy import select
 
-from core.agent.entities import AgentEntity, AgentToolEntity
+from core.agent.entities import AgentEntity, AgentToolEntity, ExecutionContext
 from core.app.app_config.features.file_upload.manager import FileUploadConfigManager
 from core.app.apps.agent_chat.app_config_manager import AgentChatAppConfig
 from core.app.apps.base_app_queue_manager import AppQueueManager
@@ -116,9 +116,20 @@ class BaseAgentRunner(AppRunner):
         features = model_schema.features if model_schema and model_schema.features else []
         self.stream_tool_call = ModelFeature.STREAM_TOOL_CALL in features
         self.files = application_generate_entity.files if ModelFeature.VISION in features else []
+        self.model_features = features
         self.query: str | None = ""
         self._current_thoughts: list[PromptMessage] = []
 
+    def build_execution_context(self) -> ExecutionContext:
+        """Build execution context."""
+        return ExecutionContext(
+            user_id=self.user_id,
+            app_id=self.app_config.app_id,
+            conversation_id=self.conversation.id,
+            message_id=self.message.id,
+            tenant_id=self.tenant_id,
+        )
+
     def _repack_app_generate_entity(
         self, app_generate_entity: AgentChatAppGenerateEntity
     ) -> AgentChatAppGenerateEntity:

api/core/agent/cot_agent_runner.py

@@ -1,437 +0,0 @@
-import json
-import logging
-from abc import ABC, abstractmethod
-from collections.abc import Generator, Mapping, Sequence
-from typing import Any
-
-from core.agent.base_agent_runner import BaseAgentRunner
-from core.agent.entities import AgentScratchpadUnit
-from core.agent.output_parser.cot_output_parser import CotAgentOutputParser
-from core.app.apps.base_app_queue_manager import PublishFrom
-from core.app.entities.queue_entities import QueueAgentThoughtEvent, QueueMessageEndEvent, QueueMessageFileEvent
-from core.model_runtime.entities.llm_entities import LLMResult, LLMResultChunk, LLMResultChunkDelta, LLMUsage
-from core.model_runtime.entities.message_entities import (
-    AssistantPromptMessage,
-    PromptMessage,
-    PromptMessageTool,
-    ToolPromptMessage,
-    UserPromptMessage,
-)
-from core.ops.ops_trace_manager import TraceQueueManager
-from core.prompt.agent_history_prompt_transform import AgentHistoryPromptTransform
-from core.tools.__base.tool import Tool
-from core.tools.entities.tool_entities import ToolInvokeMeta
-from core.tools.tool_engine import ToolEngine
-from core.workflow.nodes.agent.exc import AgentMaxIterationError
-from models.model import Message
-
-logger = logging.getLogger(__name__)
-
-
-class CotAgentRunner(BaseAgentRunner, ABC):
-    _is_first_iteration = True
-    _ignore_observation_providers = ["wenxin"]
-    _historic_prompt_messages: list[PromptMessage]
-    _agent_scratchpad: list[AgentScratchpadUnit]
-    _instruction: str
-    _query: str
-    _prompt_messages_tools: Sequence[PromptMessageTool]
-
-    def run(
-        self,
-        message: Message,
-        query: str,
-        inputs: Mapping[str, str],
-    ) -> Generator:
-        """
-        Run Cot agent application
-        """
-
-        app_generate_entity = self.application_generate_entity
-        self._repack_app_generate_entity(app_generate_entity)
-        self._init_react_state(query)
-
-        trace_manager = app_generate_entity.trace_manager
-
-        # check model mode
-        if "Observation" not in app_generate_entity.model_conf.stop:
-            if app_generate_entity.model_conf.provider not in self._ignore_observation_providers:
-                app_generate_entity.model_conf.stop.append("Observation")
-
-        app_config = self.app_config
-        assert app_config.agent
-
-        # init instruction
-        inputs = inputs or {}
-        instruction = app_config.prompt_template.simple_prompt_template or ""
-        self._instruction = self._fill_in_inputs_from_external_data_tools(instruction, inputs)
-
-        iteration_step = 1
-        max_iteration_steps = min(app_config.agent.max_iteration, 99) + 1
-
-        # convert tools into ModelRuntime Tool format
-        tool_instances, prompt_messages_tools = self._init_prompt_tools()
-        self._prompt_messages_tools = prompt_messages_tools
-
-        function_call_state = True
-        llm_usage: dict[str, LLMUsage | None] = {"usage": None}
-        final_answer = ""
-        prompt_messages: list = []  # Initialize prompt_messages
-        agent_thought_id = ""  # Initialize agent_thought_id
-
-        def increase_usage(final_llm_usage_dict: dict[str, LLMUsage | None], usage: LLMUsage):
-            if not final_llm_usage_dict["usage"]:
-                final_llm_usage_dict["usage"] = usage
-            else:
-                llm_usage = final_llm_usage_dict["usage"]
-                llm_usage.prompt_tokens += usage.prompt_tokens
-                llm_usage.completion_tokens += usage.completion_tokens
-                llm_usage.total_tokens += usage.total_tokens
-                llm_usage.prompt_price += usage.prompt_price
-                llm_usage.completion_price += usage.completion_price
-                llm_usage.total_price += usage.total_price
-
-        model_instance = self.model_instance
-
-        while function_call_state and iteration_step <= max_iteration_steps:
-            # continue to run until there is not any tool call
-            function_call_state = False
-
-            if iteration_step == max_iteration_steps:
-                # the last iteration, remove all tools
-                self._prompt_messages_tools = []
-
-            message_file_ids: list[str] = []
-
-            agent_thought_id = self.create_agent_thought(
-                message_id=message.id, message="", tool_name="", tool_input="", messages_ids=message_file_ids
-            )
-
-            if iteration_step > 1:
-                self.queue_manager.publish(
-                    QueueAgentThoughtEvent(agent_thought_id=agent_thought_id), PublishFrom.APPLICATION_MANAGER
-                )
-
-            # recalc llm max tokens
-            prompt_messages = self._organize_prompt_messages()
-            self.recalc_llm_max_tokens(self.model_config, prompt_messages)
-            # invoke model
-            chunks = model_instance.invoke_llm(
-                prompt_messages=prompt_messages,
-                model_parameters=app_generate_entity.model_conf.parameters,
-                tools=[],
-                stop=app_generate_entity.model_conf.stop,
-                stream=True,
-                user=self.user_id,
-                callbacks=[],
-            )
-
-            usage_dict: dict[str, LLMUsage | None] = {}
-            react_chunks = CotAgentOutputParser.handle_react_stream_output(chunks, usage_dict)
-            scratchpad = AgentScratchpadUnit(
-                agent_response="",
-                thought="",
-                action_str="",
-                observation="",
-                action=None,
-            )
-
-            # publish agent thought if it's first iteration
-            if iteration_step == 1:
-                self.queue_manager.publish(
-                    QueueAgentThoughtEvent(agent_thought_id=agent_thought_id), PublishFrom.APPLICATION_MANAGER
-                )
-
-            for chunk in react_chunks:
-                if isinstance(chunk, AgentScratchpadUnit.Action):
-                    action = chunk
-                    # detect action
-                    assert scratchpad.agent_response is not None
-                    scratchpad.agent_response += json.dumps(chunk.model_dump())
-                    scratchpad.action_str = json.dumps(chunk.model_dump())
-                    scratchpad.action = action
-                else:
-                    assert scratchpad.agent_response is not None
-                    scratchpad.agent_response += chunk
-                    assert scratchpad.thought is not None
-                    scratchpad.thought += chunk
-                    yield LLMResultChunk(
-                        model=self.model_config.model,
-                        prompt_messages=prompt_messages,
-                        system_fingerprint="",
-                        delta=LLMResultChunkDelta(index=0, message=AssistantPromptMessage(content=chunk), usage=None),
-                    )
-
-            assert scratchpad.thought is not None
-            scratchpad.thought = scratchpad.thought.strip() or "I am thinking about how to help you"
-            self._agent_scratchpad.append(scratchpad)
-
-            # Check if max iteration is reached and model still wants to call tools
-            if iteration_step == max_iteration_steps and scratchpad.action:
-                if scratchpad.action.action_name.lower() != "final answer":
-                    raise AgentMaxIterationError(app_config.agent.max_iteration)
-
-            # get llm usage
-            if "usage" in usage_dict:
-                if usage_dict["usage"] is not None:
-                    increase_usage(llm_usage, usage_dict["usage"])
-            else:
-                usage_dict["usage"] = LLMUsage.empty_usage()
-
-            self.save_agent_thought(
-                agent_thought_id=agent_thought_id,
-                tool_name=(scratchpad.action.action_name if scratchpad.action and not scratchpad.is_final() else ""),
-                tool_input={scratchpad.action.action_name: scratchpad.action.action_input} if scratchpad.action else {},
-                tool_invoke_meta={},
-                thought=scratchpad.thought or "",
-                observation="",
-                answer=scratchpad.agent_response or "",
-                messages_ids=[],
-                llm_usage=usage_dict["usage"],
-            )
-
-            if not scratchpad.is_final():
-                self.queue_manager.publish(
-                    QueueAgentThoughtEvent(agent_thought_id=agent_thought_id), PublishFrom.APPLICATION_MANAGER
-                )
-
-            if not scratchpad.action:
-                # failed to extract action, return final answer directly
-                final_answer = ""
-            else:
-                if scratchpad.action.action_name.lower() == "final answer":
-                    # action is final answer, return final answer directly
-                    try:
-                        if isinstance(scratchpad.action.action_input, dict):
-                            final_answer = json.dumps(scratchpad.action.action_input, ensure_ascii=False)
-                        elif isinstance(scratchpad.action.action_input, str):
-                            final_answer = scratchpad.action.action_input
-                        else:
-                            final_answer = f"{scratchpad.action.action_input}"
-                    except TypeError:
-                        final_answer = f"{scratchpad.action.action_input}"
-                else:
-                    function_call_state = True
-                    # action is tool call, invoke tool
-                    tool_invoke_response, tool_invoke_meta = self._handle_invoke_action(
-                        action=scratchpad.action,
-                        tool_instances=tool_instances,
-                        message_file_ids=message_file_ids,
-                        trace_manager=trace_manager,
-                    )
-                    scratchpad.observation = tool_invoke_response
-                    scratchpad.agent_response = tool_invoke_response
-
-                    self.save_agent_thought(
-                        agent_thought_id=agent_thought_id,
-                        tool_name=scratchpad.action.action_name,
-                        tool_input={scratchpad.action.action_name: scratchpad.action.action_input},
-                        thought=scratchpad.thought or "",
-                        observation={scratchpad.action.action_name: tool_invoke_response},
-                        tool_invoke_meta={scratchpad.action.action_name: tool_invoke_meta.to_dict()},
-                        answer=scratchpad.agent_response,
-                        messages_ids=message_file_ids,
-                        llm_usage=usage_dict["usage"],
-                    )
-
-                    self.queue_manager.publish(
-                        QueueAgentThoughtEvent(agent_thought_id=agent_thought_id), PublishFrom.APPLICATION_MANAGER
-                    )
-
-                # update prompt tool message
-                for prompt_tool in self._prompt_messages_tools:
-                    self.update_prompt_message_tool(tool_instances[prompt_tool.name], prompt_tool)
-
-            iteration_step += 1
-
-        yield LLMResultChunk(
-            model=model_instance.model,
-            prompt_messages=prompt_messages,
-            delta=LLMResultChunkDelta(
-                index=0, message=AssistantPromptMessage(content=final_answer), usage=llm_usage["usage"]
-            ),
-            system_fingerprint="",
-        )
-
-        # save agent thought
-        self.save_agent_thought(
-            agent_thought_id=agent_thought_id,
-            tool_name="",
-            tool_input={},
-            tool_invoke_meta={},
-            thought=final_answer,
-            observation={},
-            answer=final_answer,
-            messages_ids=[],
-        )
-        # publish end event
-        self.queue_manager.publish(
-            QueueMessageEndEvent(
-                llm_result=LLMResult(
-                    model=model_instance.model,
-                    prompt_messages=prompt_messages,
-                    message=AssistantPromptMessage(content=final_answer),
-                    usage=llm_usage["usage"] or LLMUsage.empty_usage(),
-                    system_fingerprint="",
-                )
-            ),
-            PublishFrom.APPLICATION_MANAGER,
-        )
-
-    def _handle_invoke_action(
-        self,
-        action: AgentScratchpadUnit.Action,
-        tool_instances: Mapping[str, Tool],
-        message_file_ids: list[str],
-        trace_manager: TraceQueueManager | None = None,
-    ) -> tuple[str, ToolInvokeMeta]:
-        """
-        handle invoke action
-        :param action: action
-        :param tool_instances: tool instances
-        :param message_file_ids: message file ids
-        :param trace_manager: trace manager
-        :return: observation, meta
-        """
-        # action is tool call, invoke tool
-        tool_call_name = action.action_name
-        tool_call_args = action.action_input
-        tool_instance = tool_instances.get(tool_call_name)
-
-        if not tool_instance:
-            answer = f"there is not a tool named {tool_call_name}"
-            return answer, ToolInvokeMeta.error_instance(answer)
-
-        if isinstance(tool_call_args, str):
-            try:
-                tool_call_args = json.loads(tool_call_args)
-            except json.JSONDecodeError:
-                pass
-
-        # invoke tool
-        tool_invoke_response, message_files, tool_invoke_meta = ToolEngine.agent_invoke(
-            tool=tool_instance,
-            tool_parameters=tool_call_args,
-            user_id=self.user_id,
-            tenant_id=self.tenant_id,
-            message=self.message,
-            invoke_from=self.application_generate_entity.invoke_from,
-            agent_tool_callback=self.agent_callback,
-            trace_manager=trace_manager,
-        )
-
-        # publish files
-        for message_file_id in message_files:
-            # publish message file
-            self.queue_manager.publish(
-                QueueMessageFileEvent(message_file_id=message_file_id), PublishFrom.APPLICATION_MANAGER
-            )
-            # add message file ids
-            message_file_ids.append(message_file_id)
-
-        return tool_invoke_response, tool_invoke_meta
-
-    def _convert_dict_to_action(self, action: dict) -> AgentScratchpadUnit.Action:
-        """
-        convert dict to action
-        """
-        return AgentScratchpadUnit.Action(action_name=action["action"], action_input=action["action_input"])
-
-    def _fill_in_inputs_from_external_data_tools(self, instruction: str, inputs: Mapping[str, Any]) -> str:
-        """
-        fill in inputs from external data tools
-        """
-        for key, value in inputs.items():
-            try:
-                instruction = instruction.replace(f"{{{{{key}}}}}", str(value))
-            except Exception:
-                continue
-
-        return instruction
-
-    def _init_react_state(self, query):
-        """
-        init agent scratchpad
-        """
-        self._query = query
-        self._agent_scratchpad = []
-        self._historic_prompt_messages = self._organize_historic_prompt_messages()
-
-    @abstractmethod
-    def _organize_prompt_messages(self) -> list[PromptMessage]:
-        """
-        organize prompt messages
-        """
-
-    def _format_assistant_message(self, agent_scratchpad: list[AgentScratchpadUnit]) -> str:
-        """
-        format assistant message
-        """
-        message = ""
-        for scratchpad in agent_scratchpad:
-            if scratchpad.is_final():
-                message += f"Final Answer: {scratchpad.agent_response}"
-            else:
-                message += f"Thought: {scratchpad.thought}\n\n"
-                if scratchpad.action_str:
-                    message += f"Action: {scratchpad.action_str}\n\n"
-                if scratchpad.observation:
-                    message += f"Observation: {scratchpad.observation}\n\n"
-
-        return message
-
-    def _organize_historic_prompt_messages(
-        self, current_session_messages: list[PromptMessage] | None = None
-    ) -> list[PromptMessage]:
-        """
-        organize historic prompt messages
-        """
-        result: list[PromptMessage] = []
-        scratchpads: list[AgentScratchpadUnit] = []
-        current_scratchpad: AgentScratchpadUnit | None = None
-
-        for message in self.history_prompt_messages:
-            if isinstance(message, AssistantPromptMessage):
-                if not current_scratchpad:
-                    assert isinstance(message.content, str)
-                    current_scratchpad = AgentScratchpadUnit(
-                        agent_response=message.content,
-                        thought=message.content or "I am thinking about how to help you",
-                        action_str="",
-                        action=None,
-                        observation=None,
-                    )
-                    scratchpads.append(current_scratchpad)
-                if message.tool_calls:
-                    try:
-                        current_scratchpad.action = AgentScratchpadUnit.Action(
-                            action_name=message.tool_calls[0].function.name,
-                            action_input=json.loads(message.tool_calls[0].function.arguments),
-                        )
-                        current_scratchpad.action_str = json.dumps(current_scratchpad.action.to_dict())
-                    except Exception:
-                        logger.exception("Failed to parse tool call from assistant message")
-            elif isinstance(message, ToolPromptMessage):
-                if current_scratchpad:
-                    assert isinstance(message.content, str)
-                    current_scratchpad.observation = message.content
-                else:
-                    raise NotImplementedError("expected str type")
-            elif isinstance(message, UserPromptMessage):
-                if scratchpads:
-                    result.append(AssistantPromptMessage(content=self._format_assistant_message(scratchpads)))
-                    scratchpads = []
-                    current_scratchpad = None
-
-                result.append(message)
-
-        if scratchpads:
-            result.append(AssistantPromptMessage(content=self._format_assistant_message(scratchpads)))
-
-        historic_prompts = AgentHistoryPromptTransform(
-            model_config=self.model_config,
-            prompt_messages=current_session_messages or [],
-            history_messages=result,
-            memory=self.memory,
-        ).get_prompt()
-        return historic_prompts

api/core/agent/cot_chat_agent_runner.py

@@ -1,118 +0,0 @@
-import json
-
-from core.agent.cot_agent_runner import CotAgentRunner
-from core.file import file_manager
-from core.model_runtime.entities import (
-    AssistantPromptMessage,
-    PromptMessage,
-    SystemPromptMessage,
-    TextPromptMessageContent,
-    UserPromptMessage,
-)
-from core.model_runtime.entities.message_entities import ImagePromptMessageContent, PromptMessageContentUnionTypes
-from core.model_runtime.utils.encoders import jsonable_encoder
-
-
-class CotChatAgentRunner(CotAgentRunner):
-    def _organize_system_prompt(self) -> SystemPromptMessage:
-        """
-        Organize system prompt
-        """
-        assert self.app_config.agent
-        assert self.app_config.agent.prompt
-
-        prompt_entity = self.app_config.agent.prompt
-        if not prompt_entity:
-            raise ValueError("Agent prompt configuration is not set")
-        first_prompt = prompt_entity.first_prompt
-
-        system_prompt = (
-            first_prompt.replace("{{instruction}}", self._instruction)
-            .replace("{{tools}}", json.dumps(jsonable_encoder(self._prompt_messages_tools)))
-            .replace("{{tool_names}}", ", ".join([tool.name for tool in self._prompt_messages_tools]))
-        )
-
-        return SystemPromptMessage(content=system_prompt)
-
-    def _organize_user_query(self, query, prompt_messages: list[PromptMessage]) -> list[PromptMessage]:
-        """
-        Organize user query
-        """
-        if self.files:
-            # get image detail config
-            image_detail_config = (
-                self.application_generate_entity.file_upload_config.image_config.detail
-                if (
-                    self.application_generate_entity.file_upload_config
-                    and self.application_generate_entity.file_upload_config.image_config
-                )
-                else None
-            )
-            image_detail_config = image_detail_config or ImagePromptMessageContent.DETAIL.LOW
-
-            prompt_message_contents: list[PromptMessageContentUnionTypes] = []
-            for file in self.files:
-                prompt_message_contents.append(
-                    file_manager.to_prompt_message_content(
-                        file,
-                        image_detail_config=image_detail_config,
-                    )
-                )
-            prompt_message_contents.append(TextPromptMessageContent(data=query))
-
-            prompt_messages.append(UserPromptMessage(content=prompt_message_contents))
-        else:
-            prompt_messages.append(UserPromptMessage(content=query))
-
-        return prompt_messages
-
-    def _organize_prompt_messages(self) -> list[PromptMessage]:
-        """
-        Organize
-        """
-        # organize system prompt
-        system_message = self._organize_system_prompt()
-
-        # organize current assistant messages
-        agent_scratchpad = self._agent_scratchpad
-        if not agent_scratchpad:
-            assistant_messages = []
-        else:
-            assistant_message = AssistantPromptMessage(content="")
-            assistant_message.content = ""  # FIXME: type check tell mypy that assistant_message.content is str
-            for unit in agent_scratchpad:
-                if unit.is_final():
-                    assert isinstance(assistant_message.content, str)
-                    assistant_message.content += f"Final Answer: {unit.agent_response}"
-                else:
-                    assert isinstance(assistant_message.content, str)
-                    assistant_message.content += f"Thought: {unit.thought}\n\n"
-                    if unit.action_str:
-                        assistant_message.content += f"Action: {unit.action_str}\n\n"
-                    if unit.observation:
-                        assistant_message.content += f"Observation: {unit.observation}\n\n"
-
-            assistant_messages = [assistant_message]
-
-        # query messages
-        query_messages = self._organize_user_query(self._query, [])
-
-        if assistant_messages:
-            # organize historic prompt messages
-            historic_messages = self._organize_historic_prompt_messages(
-                [system_message, *query_messages, *assistant_messages, UserPromptMessage(content="continue")]
-            )
-            messages = [
-                system_message,
-                *historic_messages,
-                *query_messages,
-                *assistant_messages,
-                UserPromptMessage(content="continue"),
-            ]
-        else:
-            # organize historic prompt messages
-            historic_messages = self._organize_historic_prompt_messages([system_message, *query_messages])
-            messages = [system_message, *historic_messages, *query_messages]
-
-        # join all messages
-        return messages

api/core/agent/cot_completion_agent_runner.py

@@ -1,87 +0,0 @@
-import json
-
-from core.agent.cot_agent_runner import CotAgentRunner
-from core.model_runtime.entities.message_entities import (
-    AssistantPromptMessage,
-    PromptMessage,
-    TextPromptMessageContent,
-    UserPromptMessage,
-)
-from core.model_runtime.utils.encoders import jsonable_encoder
-
-
-class CotCompletionAgentRunner(CotAgentRunner):
-    def _organize_instruction_prompt(self) -> str:
-        """
-        Organize instruction prompt
-        """
-        if self.app_config.agent is None:
-            raise ValueError("Agent configuration is not set")
-        prompt_entity = self.app_config.agent.prompt
-        if prompt_entity is None:
-            raise ValueError("prompt entity is not set")
-        first_prompt = prompt_entity.first_prompt
-
-        system_prompt = (
-            first_prompt.replace("{{instruction}}", self._instruction)
-            .replace("{{tools}}", json.dumps(jsonable_encoder(self._prompt_messages_tools)))
-            .replace("{{tool_names}}", ", ".join([tool.name for tool in self._prompt_messages_tools]))
-        )
-
-        return system_prompt
-
-    def _organize_historic_prompt(self, current_session_messages: list[PromptMessage] | None = None) -> str:
-        """
-        Organize historic prompt
-        """
-        historic_prompt_messages = self._organize_historic_prompt_messages(current_session_messages)
-        historic_prompt = ""
-
-        for message in historic_prompt_messages:
-            if isinstance(message, UserPromptMessage):
-                historic_prompt += f"Question: {message.content}\n\n"
-            elif isinstance(message, AssistantPromptMessage):
-                if isinstance(message.content, str):
-                    historic_prompt += message.content + "\n\n"
-                elif isinstance(message.content, list):
-                    for content in message.content:
-                        if not isinstance(content, TextPromptMessageContent):
-                            continue
-                        historic_prompt += content.data
-
-        return historic_prompt
-
-    def _organize_prompt_messages(self) -> list[PromptMessage]:
-        """
-        Organize prompt messages
-        """
-        # organize system prompt
-        system_prompt = self._organize_instruction_prompt()
-
-        # organize historic prompt messages
-        historic_prompt = self._organize_historic_prompt()
-
-        # organize current assistant messages
-        agent_scratchpad = self._agent_scratchpad
-        assistant_prompt = ""
-        for unit in agent_scratchpad or []:
-            if unit.is_final():
-                assistant_prompt += f"Final Answer: {unit.agent_response}"
-            else:
-                assistant_prompt += f"Thought: {unit.thought}\n\n"
-                if unit.action_str:
-                    assistant_prompt += f"Action: {unit.action_str}\n\n"
-                if unit.observation:
-                    assistant_prompt += f"Observation: {unit.observation}\n\n"
-
-        # query messages
-        query_prompt = f"Question: {self._query}"
-
-        # join all messages
-        prompt = (
-            system_prompt.replace("{{historic_messages}}", historic_prompt)
-            .replace("{{agent_scratchpad}}", assistant_prompt)
-            .replace("{{query}}", query_prompt)
-        )
-
-        return [UserPromptMessage(content=prompt)]

api/core/agent/entities.py

@@ -1,3 +1,5 @@
+import uuid
+from collections.abc import Mapping
 from enum import StrEnum
 from typing import Any, Union
 
@@ -92,3 +94,96 @@ class AgentInvokeMessage(ToolInvokeMessage):
     """
 
     pass
+
+
+class ExecutionContext(BaseModel):
+    """Execution context containing trace and audit information.
+
+    This context carries all the IDs and metadata that are not part of
+    the core business logic but needed for tracing, auditing, and
+    correlation purposes.
+    """
+
+    user_id: str | None = None
+    app_id: str | None = None
+    conversation_id: str | None = None
+    message_id: str | None = None
+    tenant_id: str | None = None
+
+    @classmethod
+    def create_minimal(cls, user_id: str | None = None) -> "ExecutionContext":
+        """Create a minimal context with only essential fields."""
+        return cls(user_id=user_id)
+
+    def to_dict(self) -> dict[str, Any]:
+        """Convert to dictionary for passing to legacy code."""
+        return {
+            "user_id": self.user_id,
+            "app_id": self.app_id,
+            "conversation_id": self.conversation_id,
+            "message_id": self.message_id,
+            "tenant_id": self.tenant_id,
+        }
+
+    def with_updates(self, **kwargs) -> "ExecutionContext":
+        """Create a new context with updated fields."""
+        data = self.to_dict()
+        data.update(kwargs)
+
+        return ExecutionContext(
+            user_id=data.get("user_id"),
+            app_id=data.get("app_id"),
+            conversation_id=data.get("conversation_id"),
+            message_id=data.get("message_id"),
+            tenant_id=data.get("tenant_id"),
+        )
+
+
+class AgentLog(BaseModel):
+    """
+    Agent Log.
+    """
+
+    class LogType(StrEnum):
+        """Type of agent log entry."""
+
+        ROUND = "round"  # A complete iteration round
+        THOUGHT = "thought"  # LLM thinking/reasoning
+        TOOL_CALL = "tool_call"  # Tool invocation
+
+    class LogMetadata(StrEnum):
+        STARTED_AT = "started_at"
+        FINISHED_AT = "finished_at"
+        ELAPSED_TIME = "elapsed_time"
+        TOTAL_PRICE = "total_price"
+        TOTAL_TOKENS = "total_tokens"
+        PROVIDER = "provider"
+        CURRENCY = "currency"
+        LLM_USAGE = "llm_usage"
+        ICON = "icon"
+        ICON_DARK = "icon_dark"
+
+    class LogStatus(StrEnum):
+        START = "start"
+        ERROR = "error"
+        SUCCESS = "success"
+
+    id: str = Field(default_factory=lambda: str(uuid.uuid4()), description="The id of the log")
+    label: str = Field(..., description="The label of the log")
+    log_type: LogType = Field(..., description="The type of the log")
+    parent_id: str | None = Field(default=None, description="Leave empty for root log")
+    error: str | None = Field(default=None, description="The error message")
+    status: LogStatus = Field(..., description="The status of the log")
+    data: Mapping[str, Any] = Field(..., description="Detailed log data")
+    metadata: Mapping[LogMetadata, Any] = Field(default={}, description="The metadata of the log")
+
+
+class AgentResult(BaseModel):
+    """
+    Agent execution result.
+    """
+
+    text: str = Field(default="", description="The generated text")
+    files: list[Any] = Field(default_factory=list, description="Files produced during execution")
+    usage: Any | None = Field(default=None, description="LLM usage statistics")
+    finish_reason: str | None = Field(default=None, description="Reason for completion")

api/core/agent/fc_agent_runner.py

@@ -1,468 +0,0 @@
-import json
-import logging
-from collections.abc import Generator
-from copy import deepcopy
-from typing import Any, Union
-
-from core.agent.base_agent_runner import BaseAgentRunner
-from core.app.apps.base_app_queue_manager import PublishFrom
-from core.app.entities.queue_entities import QueueAgentThoughtEvent, QueueMessageEndEvent, QueueMessageFileEvent
-from core.file import file_manager
-from core.model_runtime.entities import (
-    AssistantPromptMessage,
-    LLMResult,
-    LLMResultChunk,
-    LLMResultChunkDelta,
-    LLMUsage,
-    PromptMessage,
-    PromptMessageContentType,
-    SystemPromptMessage,
-    TextPromptMessageContent,
-    ToolPromptMessage,
-    UserPromptMessage,
-)
-from core.model_runtime.entities.message_entities import ImagePromptMessageContent, PromptMessageContentUnionTypes
-from core.prompt.agent_history_prompt_transform import AgentHistoryPromptTransform
-from core.tools.entities.tool_entities import ToolInvokeMeta
-from core.tools.tool_engine import ToolEngine
-from core.workflow.nodes.agent.exc import AgentMaxIterationError
-from models.model import Message
-
-logger = logging.getLogger(__name__)
-
-
-class FunctionCallAgentRunner(BaseAgentRunner):
-    def run(self, message: Message, query: str, **kwargs: Any) -> Generator[LLMResultChunk, None, None]:
-        """
-        Run FunctionCall agent application
-        """
-        self.query = query
-        app_generate_entity = self.application_generate_entity
-
-        app_config = self.app_config
-        assert app_config is not None, "app_config is required"
-        assert app_config.agent is not None, "app_config.agent is required"
-
-        # convert tools into ModelRuntime Tool format
-        tool_instances, prompt_messages_tools = self._init_prompt_tools()
-
-        assert app_config.agent
-
-        iteration_step = 1
-        max_iteration_steps = min(app_config.agent.max_iteration, 99) + 1
-
-        # continue to run until there is not any tool call
-        function_call_state = True
-        llm_usage: dict[str, LLMUsage | None] = {"usage": None}
-        final_answer = ""
-        prompt_messages: list = []  # Initialize prompt_messages
-
-        # get tracing instance
-        trace_manager = app_generate_entity.trace_manager
-
-        def increase_usage(final_llm_usage_dict: dict[str, LLMUsage | None], usage: LLMUsage):
-            if not final_llm_usage_dict["usage"]:
-                final_llm_usage_dict["usage"] = usage
-            else:
-                llm_usage = final_llm_usage_dict["usage"]
-                llm_usage.prompt_tokens += usage.prompt_tokens
-                llm_usage.completion_tokens += usage.completion_tokens
-                llm_usage.total_tokens += usage.total_tokens
-                llm_usage.prompt_price += usage.prompt_price
-                llm_usage.completion_price += usage.completion_price
-                llm_usage.total_price += usage.total_price
-
-        model_instance = self.model_instance
-
-        while function_call_state and iteration_step <= max_iteration_steps:
-            function_call_state = False
-
-            if iteration_step == max_iteration_steps:
-                # the last iteration, remove all tools
-                prompt_messages_tools = []
-
-            message_file_ids: list[str] = []
-            agent_thought_id = self.create_agent_thought(
-                message_id=message.id, message="", tool_name="", tool_input="", messages_ids=message_file_ids
-            )
-
-            # recalc llm max tokens
-            prompt_messages = self._organize_prompt_messages()
-            self.recalc_llm_max_tokens(self.model_config, prompt_messages)
-            # invoke model
-            chunks: Union[Generator[LLMResultChunk, None, None], LLMResult] = model_instance.invoke_llm(
-                prompt_messages=prompt_messages,
-                model_parameters=app_generate_entity.model_conf.parameters,
-                tools=prompt_messages_tools,
-                stop=app_generate_entity.model_conf.stop,
-                stream=self.stream_tool_call,
-                user=self.user_id,
-                callbacks=[],
-            )
-
-            tool_calls: list[tuple[str, str, dict[str, Any]]] = []
-
-            # save full response
-            response = ""
-
-            # save tool call names and inputs
-            tool_call_names = ""
-            tool_call_inputs = ""
-
-            current_llm_usage = None
-
-            if isinstance(chunks, Generator):
-                is_first_chunk = True
-                for chunk in chunks:
-                    if is_first_chunk:
-                        self.queue_manager.publish(
-                            QueueAgentThoughtEvent(agent_thought_id=agent_thought_id), PublishFrom.APPLICATION_MANAGER
-                        )
-                        is_first_chunk = False
-                    # check if there is any tool call
-                    if self.check_tool_calls(chunk):
-                        function_call_state = True
-                        tool_calls.extend(self.extract_tool_calls(chunk) or [])
-                        tool_call_names = ";".join([tool_call[1] for tool_call in tool_calls])
-                        try:
-                            tool_call_inputs = json.dumps(
-                                {tool_call[1]: tool_call[2] for tool_call in tool_calls}, ensure_ascii=False
-                            )
-                        except TypeError:
-                            # fallback: force ASCII to handle non-serializable objects
-                            tool_call_inputs = json.dumps({tool_call[1]: tool_call[2] for tool_call in tool_calls})
-
-                    if chunk.delta.message and chunk.delta.message.content:
-                        if isinstance(chunk.delta.message.content, list):
-                            for content in chunk.delta.message.content:
-                                response += content.data
-                        else:
-                            response += str(chunk.delta.message.content)
-
-                    if chunk.delta.usage:
-                        increase_usage(llm_usage, chunk.delta.usage)
-                        current_llm_usage = chunk.delta.usage
-
-                    yield chunk
-            else:
-                result = chunks
-                # check if there is any tool call
-                if self.check_blocking_tool_calls(result):
-                    function_call_state = True
-                    tool_calls.extend(self.extract_blocking_tool_calls(result) or [])
-                    tool_call_names = ";".join([tool_call[1] for tool_call in tool_calls])
-                    try:
-                        tool_call_inputs = json.dumps(
-                            {tool_call[1]: tool_call[2] for tool_call in tool_calls}, ensure_ascii=False
-                        )
-                    except TypeError:
-                        # fallback: force ASCII to handle non-serializable objects
-                        tool_call_inputs = json.dumps({tool_call[1]: tool_call[2] for tool_call in tool_calls})
-
-                if result.usage:
-                    increase_usage(llm_usage, result.usage)
-                    current_llm_usage = result.usage
-
-                if result.message and result.message.content:
-                    if isinstance(result.message.content, list):
-                        for content in result.message.content:
-                            response += content.data
-                    else:
-                        response += str(result.message.content)
-
-                if not result.message.content:
-                    result.message.content = ""
-
-                self.queue_manager.publish(
-                    QueueAgentThoughtEvent(agent_thought_id=agent_thought_id), PublishFrom.APPLICATION_MANAGER
-                )
-
-                yield LLMResultChunk(
-                    model=model_instance.model,
-                    prompt_messages=result.prompt_messages,
-                    system_fingerprint=result.system_fingerprint,
-                    delta=LLMResultChunkDelta(
-                        index=0,
-                        message=result.message,
-                        usage=result.usage,
-                    ),
-                )
-
-            assistant_message = AssistantPromptMessage(content=response, tool_calls=[])
-            if tool_calls:
-                assistant_message.tool_calls = [
-                    AssistantPromptMessage.ToolCall(
-                        id=tool_call[0],
-                        type="function",
-                        function=AssistantPromptMessage.ToolCall.ToolCallFunction(
-                            name=tool_call[1], arguments=json.dumps(tool_call[2], ensure_ascii=False)
-                        ),
-                    )
-                    for tool_call in tool_calls
-                ]
-
-            self._current_thoughts.append(assistant_message)
-
-            # save thought
-            self.save_agent_thought(
-                agent_thought_id=agent_thought_id,
-                tool_name=tool_call_names,
-                tool_input=tool_call_inputs,
-                thought=response,
-                tool_invoke_meta=None,
-                observation=None,
-                answer=response,
-                messages_ids=[],
-                llm_usage=current_llm_usage,
-            )
-            self.queue_manager.publish(
-                QueueAgentThoughtEvent(agent_thought_id=agent_thought_id), PublishFrom.APPLICATION_MANAGER
-            )
-
-            final_answer += response + "\n"
-
-            # Check if max iteration is reached and model still wants to call tools
-            if iteration_step == max_iteration_steps and tool_calls:
-                raise AgentMaxIterationError(app_config.agent.max_iteration)
-
-            # call tools
-            tool_responses = []
-            for tool_call_id, tool_call_name, tool_call_args in tool_calls:
-                tool_instance = tool_instances.get(tool_call_name)
-                if not tool_instance:
-                    tool_response = {
-                        "tool_call_id": tool_call_id,
-                        "tool_call_name": tool_call_name,
-                        "tool_response": f"there is not a tool named {tool_call_name}",
-                        "meta": ToolInvokeMeta.error_instance(f"there is not a tool named {tool_call_name}").to_dict(),
-                    }
-                else:
-                    # invoke tool
-                    tool_invoke_response, message_files, tool_invoke_meta = ToolEngine.agent_invoke(
-                        tool=tool_instance,
-                        tool_parameters=tool_call_args,
-                        user_id=self.user_id,
-                        tenant_id=self.tenant_id,
-                        message=self.message,
-                        invoke_from=self.application_generate_entity.invoke_from,
-                        agent_tool_callback=self.agent_callback,
-                        trace_manager=trace_manager,
-                        app_id=self.application_generate_entity.app_config.app_id,
-                        message_id=self.message.id,
-                        conversation_id=self.conversation.id,
-                    )
-                    # publish files
-                    for message_file_id in message_files:
-                        # publish message file
-                        self.queue_manager.publish(
-                            QueueMessageFileEvent(message_file_id=message_file_id), PublishFrom.APPLICATION_MANAGER
-                        )
-                        # add message file ids
-                        message_file_ids.append(message_file_id)
-
-                    tool_response = {
-                        "tool_call_id": tool_call_id,
-                        "tool_call_name": tool_call_name,
-                        "tool_response": tool_invoke_response,
-                        "meta": tool_invoke_meta.to_dict(),
-                    }
-
-                tool_responses.append(tool_response)
-                if tool_response["tool_response"] is not None:
-                    self._current_thoughts.append(
-                        ToolPromptMessage(
-                            content=str(tool_response["tool_response"]),
-                            tool_call_id=tool_call_id,
-                            name=tool_call_name,
-                        )
-                    )
-
-            if len(tool_responses) > 0:
-                # save agent thought
-                self.save_agent_thought(
-                    agent_thought_id=agent_thought_id,
-                    tool_name="",
-                    tool_input="",
-                    thought="",
-                    tool_invoke_meta={
-                        tool_response["tool_call_name"]: tool_response["meta"] for tool_response in tool_responses
-                    },
-                    observation={
-                        tool_response["tool_call_name"]: tool_response["tool_response"]
-                        for tool_response in tool_responses
-                    },
-                    answer="",
-                    messages_ids=message_file_ids,
-                )
-                self.queue_manager.publish(
-                    QueueAgentThoughtEvent(agent_thought_id=agent_thought_id), PublishFrom.APPLICATION_MANAGER
-                )
-
-            # update prompt tool
-            for prompt_tool in prompt_messages_tools:
-                self.update_prompt_message_tool(tool_instances[prompt_tool.name], prompt_tool)
-
-            iteration_step += 1
-
-        # publish end event
-        self.queue_manager.publish(
-            QueueMessageEndEvent(
-                llm_result=LLMResult(
-                    model=model_instance.model,
-                    prompt_messages=prompt_messages,
-                    message=AssistantPromptMessage(content=final_answer),
-                    usage=llm_usage["usage"] or LLMUsage.empty_usage(),
-                    system_fingerprint="",
-                )
-            ),
-            PublishFrom.APPLICATION_MANAGER,
-        )
-
-    def check_tool_calls(self, llm_result_chunk: LLMResultChunk) -> bool:
-        """
-        Check if there is any tool call in llm result chunk
-        """
-        if llm_result_chunk.delta.message.tool_calls:
-            return True
-        return False
-
-    def check_blocking_tool_calls(self, llm_result: LLMResult) -> bool:
-        """
-        Check if there is any blocking tool call in llm result
-        """
-        if llm_result.message.tool_calls:
-            return True
-        return False
-
-    def extract_tool_calls(self, llm_result_chunk: LLMResultChunk) -> list[tuple[str, str, dict[str, Any]]]:
-        """
-        Extract tool calls from llm result chunk
-
-        Returns:
-            List[Tuple[str, str, Dict[str, Any]]]: [(tool_call_id, tool_call_name, tool_call_args)]
-        """
-        tool_calls = []
-        for prompt_message in llm_result_chunk.delta.message.tool_calls:
-            args = {}
-            if prompt_message.function.arguments != "":
-                args = json.loads(prompt_message.function.arguments)
-
-            tool_calls.append(
-                (
-                    prompt_message.id,
-                    prompt_message.function.name,
-                    args,
-                )
-            )
-
-        return tool_calls
-
-    def extract_blocking_tool_calls(self, llm_result: LLMResult) -> list[tuple[str, str, dict[str, Any]]]:
-        """
-        Extract blocking tool calls from llm result
-
-        Returns:
-            List[Tuple[str, str, Dict[str, Any]]]: [(tool_call_id, tool_call_name, tool_call_args)]
-        """
-        tool_calls = []
-        for prompt_message in llm_result.message.tool_calls:
-            args = {}
-            if prompt_message.function.arguments != "":
-                args = json.loads(prompt_message.function.arguments)
-
-            tool_calls.append(
-                (
-                    prompt_message.id,
-                    prompt_message.function.name,
-                    args,
-                )
-            )
-
-        return tool_calls
-
-    def _init_system_message(self, prompt_template: str, prompt_messages: list[PromptMessage]) -> list[PromptMessage]:
-        """
-        Initialize system message
-        """
-        if not prompt_messages and prompt_template:
-            return [
-                SystemPromptMessage(content=prompt_template),
-            ]
-
-        if prompt_messages and not isinstance(prompt_messages[0], SystemPromptMessage) and prompt_template:
-            prompt_messages.insert(0, SystemPromptMessage(content=prompt_template))
-
-        return prompt_messages or []
-
-    def _organize_user_query(self, query: str, prompt_messages: list[PromptMessage]) -> list[PromptMessage]:
-        """
-        Organize user query
-        """
-        if self.files:
-            # get image detail config
-            image_detail_config = (
-                self.application_generate_entity.file_upload_config.image_config.detail
-                if (
-                    self.application_generate_entity.file_upload_config
-                    and self.application_generate_entity.file_upload_config.image_config
-                )
-                else None
-            )
-            image_detail_config = image_detail_config or ImagePromptMessageContent.DETAIL.LOW
-
-            prompt_message_contents: list[PromptMessageContentUnionTypes] = []
-            for file in self.files:
-                prompt_message_contents.append(
-                    file_manager.to_prompt_message_content(
-                        file,
-                        image_detail_config=image_detail_config,
-                    )
-                )
-            prompt_message_contents.append(TextPromptMessageContent(data=query))
-
-            prompt_messages.append(UserPromptMessage(content=prompt_message_contents))
-        else:
-            prompt_messages.append(UserPromptMessage(content=query))
-
-        return prompt_messages
-
-    def _clear_user_prompt_image_messages(self, prompt_messages: list[PromptMessage]) -> list[PromptMessage]:
-        """
-        As for now, gpt supports both fc and vision at the first iteration.
-        We need to remove the image messages from the prompt messages at the first iteration.
-        """
-        prompt_messages = deepcopy(prompt_messages)
-
-        for prompt_message in prompt_messages:
-            if isinstance(prompt_message, UserPromptMessage):
-                if isinstance(prompt_message.content, list):
-                    prompt_message.content = "\n".join(
-                        [
-                            content.data
-                            if content.type == PromptMessageContentType.TEXT
-                            else "[image]"
-                            if content.type == PromptMessageContentType.IMAGE
-                            else "[file]"
-                            for content in prompt_message.content
-                        ]
-                    )
-
-        return prompt_messages
-
-    def _organize_prompt_messages(self):
-        prompt_template = self.app_config.prompt_template.simple_prompt_template or ""
-        self.history_prompt_messages = self._init_system_message(prompt_template, self.history_prompt_messages)
-        query_prompt_messages = self._organize_user_query(self.query or "", [])
-
-        self.history_prompt_messages = AgentHistoryPromptTransform(
-            model_config=self.model_config,
-            prompt_messages=[*query_prompt_messages, *self._current_thoughts],
-            history_messages=self.history_prompt_messages,
-            memory=self.memory,
-        ).get_prompt()
-
-        prompt_messages = [*self.history_prompt_messages, *query_prompt_messages, *self._current_thoughts]
-        if len(self._current_thoughts) != 0:
-            # clear messages after the first iteration
-            prompt_messages = self._clear_user_prompt_image_messages(prompt_messages)
-        return prompt_messages

api/core/agent/patterns/README.md

@@ -0,0 +1,55 @@
+# Agent Patterns
+
+A unified agent pattern module that powers both Agent V2 workflow nodes and agent applications. Strategies share a common execution contract while adapting to model capabilities and tool availability.
+
+## Overview
+
+The module applies a strategy pattern around LLM/tool orchestration. `StrategyFactory` auto-selects the best implementation based on model features or an explicit agent strategy, and each strategy streams logs and usage consistently.
+
+## Key Features
+
+- **Dual strategies**
+  - `FunctionCallStrategy`: uses native LLM function/tool calling when the model exposes `TOOL_CALL`, `MULTI_TOOL_CALL`, or `STREAM_TOOL_CALL`.
+  - `ReActStrategy`: ReAct (reasoning + acting) flow driven by `CotAgentOutputParser`, used when function calling is unavailable or explicitly requested.
+- **Explicit or auto selection**
+  - `StrategyFactory.create_strategy` prefers an explicit `AgentEntity.Strategy` (FUNCTION_CALLING or CHAIN_OF_THOUGHT).
+  - Otherwise it falls back to function calling when tool-call features exist, or ReAct when they do not.
+- **Unified execution contract**
+  - `AgentPattern.run` yields streaming `AgentLog` entries and `LLMResultChunk` data, returning an `AgentResult` with text, files, usage, and `finish_reason`.
+  - Iterations are configurable and hard-capped at 99 rounds; the last round forces a final answer by withholding tools.
+- **Tool handling and hooks**
+  - Tools convert to `PromptMessageTool` objects before invocation.
+  - Optional `tool_invoke_hook` lets callers override tool execution (e.g., agent apps) while workflow runs use `ToolEngine.generic_invoke`.
+  - Tool outputs support text, links, JSON, variables, blobs, retriever resources, and file attachments; `target=="self"` files are reloaded into model context, others are returned as outputs.
+- **File-aware arguments**
+  - Tool args accept `[File: <id>]` or `[Files: <id1, id2>]` placeholders that resolve to `File` objects before invocation, enabling models to reference uploaded files safely.
+- **ReAct prompt shaping**
+  - System prompts replace `{{instruction}}`, `{{tools}}`, and `{{tool_names}}` placeholders.
+  - Adds `Observation` to stop sequences and appends scratchpad text so the model sees prior Thought/Action/Observation history.
+- **Observability and accounting**
+  - Standardized `AgentLog` entries for rounds, model thoughts, and tool calls, including usage aggregation (`LLMUsage`) across streaming and non-streaming paths.
+
+## Architecture
+
+```
+agent/patterns/
+├── base.py              # Shared utilities: logging, usage, tool invocation, file handling
+├── function_call.py     # Native function-calling loop with tool execution
+├── react.py             # ReAct loop with CoT parsing and scratchpad wiring
+└── strategy_factory.py  # Strategy selection by model features or explicit override
+```
+
+## Usage
+
+- For auto-selection:
+  - Call `StrategyFactory.create_strategy(model_features, model_instance, context, tools, files, ...)` and run the returned strategy with prompt messages and model params.
+- For explicit behavior:
+  - Pass `agent_strategy=AgentEntity.Strategy.FUNCTION_CALLING` to force native calls (falls back to ReAct if unsupported), or `CHAIN_OF_THOUGHT` to force ReAct.
+- Both strategies stream chunks and logs; collect the generator output until it returns an `AgentResult`.
+
+## Integration Points
+
+- **Model runtime**: delegates to `ModelInstance.invoke_llm` for both streaming and non-streaming calls.
+- **Tool system**: defaults to `ToolEngine.generic_invoke`, with `tool_invoke_hook` for custom callers.
+- **Files**: flows through `File` objects for tool inputs/outputs and model-context attachments.
+- **Execution context**: `ExecutionContext` fields (user/app/conversation/message) propagate to tool invocations and logging.

api/core/agent/patterns/__init__.py

@@ -0,0 +1,19 @@
+"""Agent patterns module.
+
+This module provides different strategies for agent execution:
+- FunctionCallStrategy: Uses native function/tool calling
+- ReActStrategy: Uses ReAct (Reasoning + Acting) approach
+- StrategyFactory: Factory for creating strategies based on model features
+"""
+
+from .base import AgentPattern
+from .function_call import FunctionCallStrategy
+from .react import ReActStrategy
+from .strategy_factory import StrategyFactory
+
+__all__ = [
+    "AgentPattern",
+    "FunctionCallStrategy",
+    "ReActStrategy",
+    "StrategyFactory",
+]

api/core/agent/patterns/base.py

@@ -0,0 +1,474 @@
+"""Base class for agent strategies."""
+
+from __future__ import annotations
+
+import json
+import re
+import time
+from abc import ABC, abstractmethod
+from collections.abc import Callable, Generator
+from typing import TYPE_CHECKING, Any
+
+from core.agent.entities import AgentLog, AgentResult, ExecutionContext
+from core.file import File
+from core.model_manager import ModelInstance
+from core.model_runtime.entities import (
+    AssistantPromptMessage,
+    LLMResult,
+    LLMResultChunk,
+    LLMResultChunkDelta,
+    PromptMessage,
+    PromptMessageTool,
+)
+from core.model_runtime.entities.llm_entities import LLMUsage
+from core.model_runtime.entities.message_entities import TextPromptMessageContent
+from core.tools.entities.tool_entities import ToolInvokeMessage, ToolInvokeMeta
+
+if TYPE_CHECKING:
+    from core.tools.__base.tool import Tool
+
+# Type alias for tool invoke hook
+# Returns: (response_content, message_file_ids, tool_invoke_meta)
+ToolInvokeHook = Callable[["Tool", dict[str, Any], str], tuple[str, list[str], ToolInvokeMeta]]
+
+
+class AgentPattern(ABC):
+    """Base class for agent execution strategies."""
+
+    def __init__(
+        self,
+        model_instance: ModelInstance,
+        tools: list[Tool],
+        context: ExecutionContext,
+        max_iterations: int = 10,
+        workflow_call_depth: int = 0,
+        files: list[File] = [],
+        tool_invoke_hook: ToolInvokeHook | None = None,
+    ):
+        """Initialize the agent strategy."""
+        self.model_instance = model_instance
+        self.tools = tools
+        self.context = context
+        self.max_iterations = min(max_iterations, 99)  # Cap at 99 iterations
+        self.workflow_call_depth = workflow_call_depth
+        self.files: list[File] = files
+        self.tool_invoke_hook = tool_invoke_hook
+
+    @abstractmethod
+    def run(
+        self,
+        prompt_messages: list[PromptMessage],
+        model_parameters: dict[str, Any],
+        stop: list[str] = [],
+        stream: bool = True,
+    ) -> Generator[LLMResultChunk | AgentLog, None, AgentResult]:
+        """Execute the agent strategy."""
+        pass
+
+    def _accumulate_usage(self, total_usage: dict[str, Any], delta_usage: LLMUsage) -> None:
+        """Accumulate LLM usage statistics."""
+        if not total_usage.get("usage"):
+            # Create a copy to avoid modifying the original
+            total_usage["usage"] = LLMUsage(
+                prompt_tokens=delta_usage.prompt_tokens,
+                prompt_unit_price=delta_usage.prompt_unit_price,
+                prompt_price_unit=delta_usage.prompt_price_unit,
+                prompt_price=delta_usage.prompt_price,
+                completion_tokens=delta_usage.completion_tokens,
+                completion_unit_price=delta_usage.completion_unit_price,
+                completion_price_unit=delta_usage.completion_price_unit,
+                completion_price=delta_usage.completion_price,
+                total_tokens=delta_usage.total_tokens,
+                total_price=delta_usage.total_price,
+                currency=delta_usage.currency,
+                latency=delta_usage.latency,
+            )
+        else:
+            current: LLMUsage = total_usage["usage"]
+            current.prompt_tokens += delta_usage.prompt_tokens
+            current.completion_tokens += delta_usage.completion_tokens
+            current.total_tokens += delta_usage.total_tokens
+            current.prompt_price += delta_usage.prompt_price
+            current.completion_price += delta_usage.completion_price
+            current.total_price += delta_usage.total_price
+
+    def _extract_content(self, content: Any) -> str:
+        """Extract text content from message content."""
+        if isinstance(content, list):
+            # Content items are PromptMessageContentUnionTypes
+            text_parts = []
+            for c in content:
+                # Check if it's a TextPromptMessageContent (which has data attribute)
+                if isinstance(c, TextPromptMessageContent):
+                    text_parts.append(c.data)
+            return "".join(text_parts)
+        return str(content)
+
+    def _has_tool_calls(self, chunk: LLMResultChunk) -> bool:
+        """Check if chunk contains tool calls."""
+        # LLMResultChunk always has delta attribute
+        return bool(chunk.delta.message and chunk.delta.message.tool_calls)
+
+    def _has_tool_calls_result(self, result: LLMResult) -> bool:
+        """Check if result contains tool calls (non-streaming)."""
+        # LLMResult always has message attribute
+        return bool(result.message and result.message.tool_calls)
+
+    def _extract_tool_calls(self, chunk: LLMResultChunk) -> list[tuple[str, str, dict[str, Any]]]:
+        """Extract tool calls from streaming chunk."""
+        tool_calls: list[tuple[str, str, dict[str, Any]]] = []
+        if chunk.delta.message and chunk.delta.message.tool_calls:
+            for tool_call in chunk.delta.message.tool_calls:
+                if tool_call.function:
+                    try:
+                        args = json.loads(tool_call.function.arguments) if tool_call.function.arguments else {}
+                    except json.JSONDecodeError:
+                        args = {}
+                    tool_calls.append((tool_call.id or "", tool_call.function.name, args))
+        return tool_calls
+
+    def _extract_tool_calls_result(self, result: LLMResult) -> list[tuple[str, str, dict[str, Any]]]:
+        """Extract tool calls from non-streaming result."""
+        tool_calls = []
+        if result.message and result.message.tool_calls:
+            for tool_call in result.message.tool_calls:
+                if tool_call.function:
+                    try:
+                        args = json.loads(tool_call.function.arguments) if tool_call.function.arguments else {}
+                    except json.JSONDecodeError:
+                        args = {}
+                    tool_calls.append((tool_call.id or "", tool_call.function.name, args))
+        return tool_calls
+
+    def _extract_text_from_message(self, message: PromptMessage) -> str:
+        """Extract text content from a prompt message."""
+        # PromptMessage always has content attribute
+        content = message.content
+        if isinstance(content, str):
+            return content
+        elif isinstance(content, list):
+            # Extract text from content list
+            text_parts = []
+            for item in content:
+                if isinstance(item, TextPromptMessageContent):
+                    text_parts.append(item.data)
+            return " ".join(text_parts)
+        return ""
+
+    def _get_tool_metadata(self, tool_instance: Tool) -> dict[AgentLog.LogMetadata, Any]:
+        """Get metadata for a tool including provider and icon info."""
+        from core.tools.tool_manager import ToolManager
+
+        metadata: dict[AgentLog.LogMetadata, Any] = {}
+        if tool_instance.entity and tool_instance.entity.identity:
+            identity = tool_instance.entity.identity
+            if identity.provider:
+                metadata[AgentLog.LogMetadata.PROVIDER] = identity.provider
+
+            # Get icon using ToolManager for proper URL generation
+            tenant_id = self.context.tenant_id
+            if tenant_id and identity.provider:
+                try:
+                    provider_type = tool_instance.tool_provider_type()
+                    icon = ToolManager.get_tool_icon(tenant_id, provider_type, identity.provider)
+                    if isinstance(icon, str):
+                        metadata[AgentLog.LogMetadata.ICON] = icon
+                    elif isinstance(icon, dict):
+                        # Handle icon dict with background/content or light/dark variants
+                        metadata[AgentLog.LogMetadata.ICON] = icon
+                except Exception:
+                    # Fallback to identity.icon if ToolManager fails
+                    if identity.icon:
+                        metadata[AgentLog.LogMetadata.ICON] = identity.icon
+            elif identity.icon:
+                metadata[AgentLog.LogMetadata.ICON] = identity.icon
+        return metadata
+
+    def _create_log(
+        self,
+        label: str,
+        log_type: AgentLog.LogType,
+        status: AgentLog.LogStatus,
+        data: dict[str, Any] | None = None,
+        parent_id: str | None = None,
+        extra_metadata: dict[AgentLog.LogMetadata, Any] | None = None,
+    ) -> AgentLog:
+        """Create a new AgentLog with standard metadata."""
+        metadata: dict[AgentLog.LogMetadata, Any] = {
+            AgentLog.LogMetadata.STARTED_AT: time.perf_counter(),
+        }
+        if extra_metadata:
+            metadata.update(extra_metadata)
+
+        return AgentLog(
+            label=label,
+            log_type=log_type,
+            status=status,
+            data=data or {},
+            parent_id=parent_id,
+            metadata=metadata,
+        )
+
+    def _finish_log(
+        self,
+        log: AgentLog,
+        data: dict[str, Any] | None = None,
+        usage: LLMUsage | None = None,
+    ) -> AgentLog:
+        """Finish an AgentLog by updating its status and metadata."""
+        log.status = AgentLog.LogStatus.SUCCESS
+
+        if data is not None:
+            log.data = data
+
+        # Calculate elapsed time
+        started_at = log.metadata.get(AgentLog.LogMetadata.STARTED_AT, time.perf_counter())
+        finished_at = time.perf_counter()
+
+        # Update metadata
+        log.metadata = {
+            **log.metadata,
+            AgentLog.LogMetadata.FINISHED_AT: finished_at,
+            # Calculate elapsed time in seconds
+            AgentLog.LogMetadata.ELAPSED_TIME: round(finished_at - started_at, 4),
+        }
+
+        # Add usage information if provided
+        if usage:
+            log.metadata.update(
+                {
+                    AgentLog.LogMetadata.TOTAL_PRICE: usage.total_price,
+                    AgentLog.LogMetadata.CURRENCY: usage.currency,
+                    AgentLog.LogMetadata.TOTAL_TOKENS: usage.total_tokens,
+                    AgentLog.LogMetadata.LLM_USAGE: usage,
+                }
+            )
+
+        return log
+
+    def _replace_file_references(self, tool_args: dict[str, Any]) -> dict[str, Any]:
+        """
+        Replace file references in tool arguments with actual File objects.
+
+        Args:
+            tool_args: Dictionary of tool arguments
+
+        Returns:
+            Updated tool arguments with file references replaced
+        """
+        # Process each argument in the dictionary
+        processed_args: dict[str, Any] = {}
+        for key, value in tool_args.items():
+            processed_args[key] = self._process_file_reference(value)
+        return processed_args
+
+    def _process_file_reference(self, data: Any) -> Any:
+        """
+        Recursively process data to replace file references.
+        Supports both single file [File: file_id] and multiple files [Files: file_id1, file_id2, ...].
+
+        Args:
+            data: The data to process (can be dict, list, str, or other types)
+
+        Returns:
+            Processed data with file references replaced
+        """
+        single_file_pattern = re.compile(r"^\[File:\s*([^\]]+)\]$")
+        multiple_files_pattern = re.compile(r"^\[Files:\s*([^\]]+)\]$")
+
+        if isinstance(data, dict):
+            # Process dictionary recursively
+            return {key: self._process_file_reference(value) for key, value in data.items()}
+        elif isinstance(data, list):
+            # Process list recursively
+            return [self._process_file_reference(item) for item in data]
+        elif isinstance(data, str):
+            # Check for single file pattern [File: file_id]
+            single_match = single_file_pattern.match(data.strip())
+            if single_match:
+                file_id = single_match.group(1).strip()
+                # Find the file in self.files
+                for file in self.files:
+                    if file.id and str(file.id) == file_id:
+                        return file
+                # If file not found, return original value
+                return data
+
+            # Check for multiple files pattern [Files: file_id1, file_id2, ...]
+            multiple_match = multiple_files_pattern.match(data.strip())
+            if multiple_match:
+                file_ids_str = multiple_match.group(1).strip()
+                # Split by comma and strip whitespace
+                file_ids = [fid.strip() for fid in file_ids_str.split(",")]
+
+                # Find all matching files
+                matched_files: list[File] = []
+                for file_id in file_ids:
+                    for file in self.files:
+                        if file.id and str(file.id) == file_id:
+                            matched_files.append(file)
+                            break
+
+                # Return list of files if any were found, otherwise return original
+                return matched_files or data
+
+            return data
+        else:
+            # Return other types as-is
+            return data
+
+    def _create_text_chunk(self, text: str, prompt_messages: list[PromptMessage]) -> LLMResultChunk:
+        """Create a text chunk for streaming."""
+        return LLMResultChunk(
+            model=self.model_instance.model,
+            prompt_messages=prompt_messages,
+            delta=LLMResultChunkDelta(
+                index=0,
+                message=AssistantPromptMessage(content=text),
+                usage=None,
+            ),
+            system_fingerprint="",
+        )
+
+    def _invoke_tool(
+        self,
+        tool_instance: Tool,
+        tool_args: dict[str, Any],
+        tool_name: str,
+    ) -> tuple[str, list[File], ToolInvokeMeta | None]:
+        """
+        Invoke a tool and collect its response.
+
+        Args:
+            tool_instance: The tool instance to invoke
+            tool_args: Tool arguments
+            tool_name: Name of the tool
+
+        Returns:
+            Tuple of (response_content, tool_files, tool_invoke_meta)
+        """
+        # Process tool_args to replace file references with actual File objects
+        tool_args = self._replace_file_references(tool_args)
+
+        # If a tool invoke hook is set, use it instead of generic_invoke
+        if self.tool_invoke_hook:
+            response_content, _, tool_invoke_meta = self.tool_invoke_hook(tool_instance, tool_args, tool_name)
+            # Note: message_file_ids are stored in DB, we don't convert them to File objects here
+            # The caller (AgentAppRunner) handles file publishing
+            return response_content, [], tool_invoke_meta
+
+        # Default: use generic_invoke for workflow scenarios
+        # Import here to avoid circular import
+        from core.tools.tool_engine import DifyWorkflowCallbackHandler, ToolEngine
+
+        tool_response = ToolEngine().generic_invoke(
+            tool=tool_instance,
+            tool_parameters=tool_args,
+            user_id=self.context.user_id or "",
+            workflow_tool_callback=DifyWorkflowCallbackHandler(),
+            workflow_call_depth=self.workflow_call_depth,
+            app_id=self.context.app_id,
+            conversation_id=self.context.conversation_id,
+            message_id=self.context.message_id,
+        )
+
+        # Collect response and files
+        response_content = ""
+        tool_files: list[File] = []
+
+        for response in tool_response:
+            if response.type == ToolInvokeMessage.MessageType.TEXT:
+                assert isinstance(response.message, ToolInvokeMessage.TextMessage)
+                response_content += response.message.text
+
+            elif response.type == ToolInvokeMessage.MessageType.LINK:
+                # Handle link messages
+                if isinstance(response.message, ToolInvokeMessage.TextMessage):
+                    response_content += f"[Link: {response.message.text}]"
+
+            elif response.type == ToolInvokeMessage.MessageType.IMAGE:
+                # Handle image URL messages
+                if isinstance(response.message, ToolInvokeMessage.TextMessage):
+                    response_content += f"[Image: {response.message.text}]"
+
+            elif response.type == ToolInvokeMessage.MessageType.IMAGE_LINK:
+                # Handle image link messages
+                if isinstance(response.message, ToolInvokeMessage.TextMessage):
+                    response_content += f"[Image: {response.message.text}]"
+
+            elif response.type == ToolInvokeMessage.MessageType.BINARY_LINK:
+                # Handle binary file link messages
+                if isinstance(response.message, ToolInvokeMessage.TextMessage):
+                    filename = response.meta.get("filename", "file") if response.meta else "file"
+                    response_content += f"[File: {filename} - {response.message.text}]"
+
+            elif response.type == ToolInvokeMessage.MessageType.JSON:
+                # Handle JSON messages
+                if isinstance(response.message, ToolInvokeMessage.JsonMessage):
+                    response_content += json.dumps(response.message.json_object, ensure_ascii=False, indent=2)
+
+            elif response.type == ToolInvokeMessage.MessageType.BLOB:
+                # Handle blob messages - convert to text representation
+                if isinstance(response.message, ToolInvokeMessage.BlobMessage):
+                    mime_type = (
+                        response.meta.get("mime_type", "application/octet-stream")
+                        if response.meta
+                        else "application/octet-stream"
+                    )
+                    size = len(response.message.blob)
+                    response_content += f"[Binary data: {mime_type}, size: {size} bytes]"
+
+            elif response.type == ToolInvokeMessage.MessageType.VARIABLE:
+                # Handle variable messages
+                if isinstance(response.message, ToolInvokeMessage.VariableMessage):
+                    var_name = response.message.variable_name
+                    var_value = response.message.variable_value
+                    if isinstance(var_value, str):
+                        response_content += var_value
+                    else:
+                        response_content += f"[Variable {var_name}: {json.dumps(var_value, ensure_ascii=False)}]"
+
+            elif response.type == ToolInvokeMessage.MessageType.BLOB_CHUNK:
+                # Handle blob chunk messages - these are parts of a larger blob
+                if isinstance(response.message, ToolInvokeMessage.BlobChunkMessage):
+                    response_content += f"[Blob chunk {response.message.sequence}: {len(response.message.blob)} bytes]"
+
+            elif response.type == ToolInvokeMessage.MessageType.RETRIEVER_RESOURCES:
+                # Handle retriever resources messages
+                if isinstance(response.message, ToolInvokeMessage.RetrieverResourceMessage):
+                    response_content += response.message.context
+
+            elif response.type == ToolInvokeMessage.MessageType.FILE:
+                # Extract file from meta
+                if response.meta and "file" in response.meta:
+                    file = response.meta["file"]
+                    if isinstance(file, File):
+                        # Check if file is for model or tool output
+                        if response.meta.get("target") == "self":
+                            # File is for model - add to files for next prompt
+                            self.files.append(file)
+                            response_content += f"File '{file.filename}' has been loaded into your context."
+                        else:
+                            # File is tool output
+                            tool_files.append(file)
+
+        return response_content, tool_files, None
+
+    def _find_tool_by_name(self, tool_name: str) -> Tool | None:
+        """Find a tool instance by its name."""
+        for tool in self.tools:
+            if tool.entity.identity.name == tool_name:
+                return tool
+        return None
+
+    def _convert_tools_to_prompt_format(self) -> list[PromptMessageTool]:
+        """Convert tools to prompt message format."""
+        prompt_tools: list[PromptMessageTool] = []
+        for tool in self.tools:
+            prompt_tools.append(tool.to_prompt_message_tool())
+        return prompt_tools
+
+    def _update_usage_with_empty(self, llm_usage: dict[str, Any]) -> None:
+        """Initialize usage tracking with empty usage if not set."""
+        if "usage" not in llm_usage or llm_usage["usage"] is None:
+            llm_usage["usage"] = LLMUsage.empty_usage()

api/core/agent/patterns/function_call.py

@@ -0,0 +1,299 @@
+"""Function Call strategy implementation."""
+
+import json
+from collections.abc import Generator
+from typing import Any, Union
+
+from core.agent.entities import AgentLog, AgentResult
+from core.file import File
+from core.model_runtime.entities import (
+    AssistantPromptMessage,
+    LLMResult,
+    LLMResultChunk,
+    LLMResultChunkDelta,
+    LLMUsage,
+    PromptMessage,
+    PromptMessageTool,
+    ToolPromptMessage,
+)
+from core.tools.entities.tool_entities import ToolInvokeMeta
+
+from .base import AgentPattern
+
+
+class FunctionCallStrategy(AgentPattern):
+    """Function Call strategy using model's native tool calling capability."""
+
+    def run(
+        self,
+        prompt_messages: list[PromptMessage],
+        model_parameters: dict[str, Any],
+        stop: list[str] = [],
+        stream: bool = True,
+    ) -> Generator[LLMResultChunk | AgentLog, None, AgentResult]:
+        """Execute the function call agent strategy."""
+        # Convert tools to prompt format
+        prompt_tools: list[PromptMessageTool] = self._convert_tools_to_prompt_format()
+
+        # Initialize tracking
+        iteration_step: int = 1
+        max_iterations: int = self.max_iterations + 1
+        function_call_state: bool = True
+        total_usage: dict[str, LLMUsage | None] = {"usage": None}
+        messages: list[PromptMessage] = list(prompt_messages)  # Create mutable copy
+        final_text: str = ""
+        finish_reason: str | None = None
+        output_files: list[File] = []  # Track files produced by tools
+
+        while function_call_state and iteration_step <= max_iterations:
+            function_call_state = False
+            round_log = self._create_log(
+                label=f"ROUND {iteration_step}",
+                log_type=AgentLog.LogType.ROUND,
+                status=AgentLog.LogStatus.START,
+                data={},
+            )
+            yield round_log
+            # On last iteration, remove tools to force final answer
+            current_tools: list[PromptMessageTool] = [] if iteration_step == max_iterations else prompt_tools
+            model_log = self._create_log(
+                label=f"{self.model_instance.model} Thought",
+                log_type=AgentLog.LogType.THOUGHT,
+                status=AgentLog.LogStatus.START,
+                data={},
+                parent_id=round_log.id,
+                extra_metadata={
+                    AgentLog.LogMetadata.PROVIDER: self.model_instance.provider,
+                },
+            )
+            yield model_log
+
+            # Track usage for this round only
+            round_usage: dict[str, LLMUsage | None] = {"usage": None}
+
+            # Invoke model
+            chunks: Union[Generator[LLMResultChunk, None, None], LLMResult] = self.model_instance.invoke_llm(
+                prompt_messages=messages,
+                model_parameters=model_parameters,
+                tools=current_tools,
+                stop=stop,
+                stream=stream,
+                user=self.context.user_id,
+                callbacks=[],
+            )
+
+            # Process response
+            tool_calls, response_content, chunk_finish_reason = yield from self._handle_chunks(
+                chunks, round_usage, model_log
+            )
+            messages.append(self._create_assistant_message(response_content, tool_calls))
+
+            # Accumulate to total usage
+            round_usage_value = round_usage.get("usage")
+            if round_usage_value:
+                self._accumulate_usage(total_usage, round_usage_value)
+
+            # Update final text if no tool calls (this is likely the final answer)
+            if not tool_calls:
+                final_text = response_content
+
+            # Update finish reason
+            if chunk_finish_reason:
+                finish_reason = chunk_finish_reason
+
+            # Process tool calls
+            tool_outputs: dict[str, str] = {}
+            if tool_calls:
+                function_call_state = True
+                # Execute tools
+                for tool_call_id, tool_name, tool_args in tool_calls:
+                    tool_response, tool_files, _ = yield from self._handle_tool_call(
+                        tool_name, tool_args, tool_call_id, messages, round_log
+                    )
+                    tool_outputs[tool_name] = tool_response
+                    # Track files produced by tools
+                    output_files.extend(tool_files)
+            yield self._finish_log(
+                round_log,
+                data={
+                    "llm_result": response_content,
+                    "tool_calls": [
+                        {"name": tc[1], "args": tc[2], "output": tool_outputs.get(tc[1], "")} for tc in tool_calls
+                    ]
+                    if tool_calls
+                    else [],
+                    "final_answer": final_text if not function_call_state else None,
+                },
+                usage=round_usage.get("usage"),
+            )
+            iteration_step += 1
+
+        # Return final result
+        from core.agent.entities import AgentResult
+
+        return AgentResult(
+            text=final_text,
+            files=output_files,
+            usage=total_usage.get("usage") or LLMUsage.empty_usage(),
+            finish_reason=finish_reason,
+        )
+
+    def _handle_chunks(
+        self,
+        chunks: Union[Generator[LLMResultChunk, None, None], LLMResult],
+        llm_usage: dict[str, LLMUsage | None],
+        start_log: AgentLog,
+    ) -> Generator[
+        LLMResultChunk | AgentLog,
+        None,
+        tuple[list[tuple[str, str, dict[str, Any]]], str, str | None],
+    ]:
+        """Handle LLM response chunks and extract tool calls and content.
+
+        Returns a tuple of (tool_calls, response_content, finish_reason).
+        """
+        tool_calls: list[tuple[str, str, dict[str, Any]]] = []
+        response_content: str = ""
+        finish_reason: str | None = None
+        if isinstance(chunks, Generator):
+            # Streaming response
+            for chunk in chunks:
+                # Extract tool calls
+                if self._has_tool_calls(chunk):
+                    tool_calls.extend(self._extract_tool_calls(chunk))
+
+                # Extract content
+                if chunk.delta.message and chunk.delta.message.content:
+                    response_content += self._extract_content(chunk.delta.message.content)
+
+                # Track usage
+                if chunk.delta.usage:
+                    self._accumulate_usage(llm_usage, chunk.delta.usage)
+
+                # Capture finish reason
+                if chunk.delta.finish_reason:
+                    finish_reason = chunk.delta.finish_reason
+
+                yield chunk
+        else:
+            # Non-streaming response
+            result: LLMResult = chunks
+
+            if self._has_tool_calls_result(result):
+                tool_calls.extend(self._extract_tool_calls_result(result))
+
+            if result.message and result.message.content:
+                response_content += self._extract_content(result.message.content)
+
+            if result.usage:
+                self._accumulate_usage(llm_usage, result.usage)
+
+            # Convert to streaming format
+            yield LLMResultChunk(
+                model=result.model,
+                prompt_messages=result.prompt_messages,
+                delta=LLMResultChunkDelta(index=0, message=result.message, usage=result.usage),
+            )
+        yield self._finish_log(
+            start_log,
+            data={
+                "result": response_content,
+            },
+            usage=llm_usage.get("usage"),
+        )
+        return tool_calls, response_content, finish_reason
+
+    def _create_assistant_message(
+        self, content: str, tool_calls: list[tuple[str, str, dict[str, Any]]] | None = None
+    ) -> AssistantPromptMessage:
+        """Create assistant message with tool calls."""
+        if tool_calls is None:
+            return AssistantPromptMessage(content=content)
+        return AssistantPromptMessage(
+            content=content or "",
+            tool_calls=[
+                AssistantPromptMessage.ToolCall(
+                    id=tc[0],
+                    type="function",
+                    function=AssistantPromptMessage.ToolCall.ToolCallFunction(name=tc[1], arguments=json.dumps(tc[2])),
+                )
+                for tc in tool_calls
+            ],
+        )
+
+    def _handle_tool_call(
+        self,
+        tool_name: str,
+        tool_args: dict[str, Any],
+        tool_call_id: str,
+        messages: list[PromptMessage],
+        round_log: AgentLog,
+    ) -> Generator[AgentLog, None, tuple[str, list[File], ToolInvokeMeta | None]]:
+        """Handle a single tool call and return response with files and meta."""
+        # Find tool
+        tool_instance = self._find_tool_by_name(tool_name)
+        if not tool_instance:
+            raise ValueError(f"Tool {tool_name} not found")
+
+        # Get tool metadata (provider, icon, etc.)
+        tool_metadata = self._get_tool_metadata(tool_instance)
+
+        # Create tool call log
+        tool_call_log = self._create_log(
+            label=f"CALL {tool_name}",
+            log_type=AgentLog.LogType.TOOL_CALL,
+            status=AgentLog.LogStatus.START,
+            data={
+                "tool_call_id": tool_call_id,
+                "tool_name": tool_name,
+                "tool_args": tool_args,
+            },
+            parent_id=round_log.id,
+            extra_metadata=tool_metadata,
+        )
+        yield tool_call_log
+
+        # Invoke tool using base class method with error handling
+        try:
+            response_content, tool_files, tool_invoke_meta = self._invoke_tool(tool_instance, tool_args, tool_name)
+
+            yield self._finish_log(
+                tool_call_log,
+                data={
+                    **tool_call_log.data,
+                    "output": response_content,
+                    "files": len(tool_files),
+                    "meta": tool_invoke_meta.to_dict() if tool_invoke_meta else None,
+                },
+            )
+            final_content = response_content or "Tool executed successfully"
+            # Add tool response to messages
+            messages.append(
+                ToolPromptMessage(
+                    content=final_content,
+                    tool_call_id=tool_call_id,
+                    name=tool_name,
+                )
+            )
+            return response_content, tool_files, tool_invoke_meta
+        except Exception as e:
+            # Tool invocation failed, yield error log
+            error_message = str(e)
+            tool_call_log.status = AgentLog.LogStatus.ERROR
+            tool_call_log.error = error_message
+            tool_call_log.data = {
+                **tool_call_log.data,
+                "error": error_message,
+            }
+            yield tool_call_log
+
+            # Add error message to conversation
+            error_content = f"Tool execution failed: {error_message}"
+            messages.append(
+                ToolPromptMessage(
+                    content=error_content,
+                    tool_call_id=tool_call_id,
+                    name=tool_name,
+                )
+            )
+            return error_content, [], None

api/core/agent/patterns/react.py

@@ -0,0 +1,418 @@
+"""ReAct strategy implementation."""
+
+from __future__ import annotations
+
+import json
+from collections.abc import Generator
+from typing import TYPE_CHECKING, Any, Union
+
+from core.agent.entities import AgentLog, AgentResult, AgentScratchpadUnit, ExecutionContext
+from core.agent.output_parser.cot_output_parser import CotAgentOutputParser
+from core.file import File
+from core.model_manager import ModelInstance
+from core.model_runtime.entities import (
+    AssistantPromptMessage,
+    LLMResult,
+    LLMResultChunk,
+    LLMResultChunkDelta,
+    PromptMessage,
+    SystemPromptMessage,
+)
+
+from .base import AgentPattern, ToolInvokeHook
+
+if TYPE_CHECKING:
+    from core.tools.__base.tool import Tool
+
+
+class ReActStrategy(AgentPattern):
+    """ReAct strategy using reasoning and acting approach."""
+
+    def __init__(
+        self,
+        model_instance: ModelInstance,
+        tools: list[Tool],
+        context: ExecutionContext,
+        max_iterations: int = 10,
+        workflow_call_depth: int = 0,
+        files: list[File] = [],
+        tool_invoke_hook: ToolInvokeHook | None = None,
+        instruction: str = "",
+    ):
+        """Initialize the ReAct strategy with instruction support."""
+        super().__init__(
+            model_instance=model_instance,
+            tools=tools,
+            context=context,
+            max_iterations=max_iterations,
+            workflow_call_depth=workflow_call_depth,
+            files=files,
+            tool_invoke_hook=tool_invoke_hook,
+        )
+        self.instruction = instruction
+
+    def run(
+        self,
+        prompt_messages: list[PromptMessage],
+        model_parameters: dict[str, Any],
+        stop: list[str] = [],
+        stream: bool = True,
+    ) -> Generator[LLMResultChunk | AgentLog, None, AgentResult]:
+        """Execute the ReAct agent strategy."""
+        # Initialize tracking
+        agent_scratchpad: list[AgentScratchpadUnit] = []
+        iteration_step: int = 1
+        max_iterations: int = self.max_iterations + 1
+        react_state: bool = True
+        total_usage: dict[str, Any] = {"usage": None}
+        output_files: list[File] = []  # Track files produced by tools
+        final_text: str = ""
+        finish_reason: str | None = None
+
+        # Add "Observation" to stop sequences
+        if "Observation" not in stop:
+            stop = stop.copy()
+            stop.append("Observation")
+
+        while react_state and iteration_step <= max_iterations:
+            react_state = False
+            round_log = self._create_log(
+                label=f"ROUND {iteration_step}",
+                log_type=AgentLog.LogType.ROUND,
+                status=AgentLog.LogStatus.START,
+                data={},
+            )
+            yield round_log
+
+            # Build prompt with/without tools based on iteration
+            include_tools = iteration_step < max_iterations
+            current_messages = self._build_prompt_with_react_format(
+                prompt_messages, agent_scratchpad, include_tools, self.instruction
+            )
+
+            model_log = self._create_log(
+                label=f"{self.model_instance.model} Thought",
+                log_type=AgentLog.LogType.THOUGHT,
+                status=AgentLog.LogStatus.START,
+                data={},
+                parent_id=round_log.id,
+                extra_metadata={
+                    AgentLog.LogMetadata.PROVIDER: self.model_instance.provider,
+                },
+            )
+            yield model_log
+
+            # Track usage for this round only
+            round_usage: dict[str, Any] = {"usage": None}
+
+            # Use current messages directly (files are handled by base class if needed)
+            messages_to_use = current_messages
+
+            # Invoke model
+            chunks: Union[Generator[LLMResultChunk, None, None], LLMResult] = self.model_instance.invoke_llm(
+                prompt_messages=messages_to_use,
+                model_parameters=model_parameters,
+                stop=stop,
+                stream=stream,
+                user=self.context.user_id or "",
+                callbacks=[],
+            )
+
+            # Process response
+            scratchpad, chunk_finish_reason = yield from self._handle_chunks(
+                chunks, round_usage, model_log, current_messages
+            )
+            agent_scratchpad.append(scratchpad)
+
+            # Accumulate to total usage
+            round_usage_value = round_usage.get("usage")
+            if round_usage_value:
+                self._accumulate_usage(total_usage, round_usage_value)
+
+            # Update finish reason
+            if chunk_finish_reason:
+                finish_reason = chunk_finish_reason
+
+            # Check if we have an action to execute
+            if scratchpad.action and scratchpad.action.action_name.lower() != "final answer":
+                react_state = True
+                # Execute tool
+                observation, tool_files = yield from self._handle_tool_call(
+                    scratchpad.action, current_messages, round_log
+                )
+                scratchpad.observation = observation
+                # Track files produced by tools
+                output_files.extend(tool_files)
+
+                # Add observation to scratchpad for display
+                yield self._create_text_chunk(f"\nObservation: {observation}\n", current_messages)
+            else:
+                # Extract final answer
+                if scratchpad.action and scratchpad.action.action_input:
+                    final_answer = scratchpad.action.action_input
+                    if isinstance(final_answer, dict):
+                        final_answer = json.dumps(final_answer, ensure_ascii=False)
+                    final_text = str(final_answer)
+                elif scratchpad.thought:
+                    # If no action but we have thought, use thought as final answer
+                    final_text = scratchpad.thought
+
+            yield self._finish_log(
+                round_log,
+                data={
+                    "thought": scratchpad.thought,
+                    "action": scratchpad.action_str if scratchpad.action else None,
+                    "observation": scratchpad.observation or None,
+                    "final_answer": final_text if not react_state else None,
+                },
+                usage=round_usage.get("usage"),
+            )
+            iteration_step += 1
+
+        # Return final result
+
+        from core.agent.entities import AgentResult
+
+        return AgentResult(
+            text=final_text, files=output_files, usage=total_usage.get("usage"), finish_reason=finish_reason
+        )
+
+    def _build_prompt_with_react_format(
+        self,
+        original_messages: list[PromptMessage],
+        agent_scratchpad: list[AgentScratchpadUnit],
+        include_tools: bool = True,
+        instruction: str = "",
+    ) -> list[PromptMessage]:
+        """Build prompt messages with ReAct format."""
+        # Copy messages to avoid modifying original
+        messages = list(original_messages)
+
+        # Find and update the system prompt that should already exist
+        system_prompt_found = False
+        for i, msg in enumerate(messages):
+            if isinstance(msg, SystemPromptMessage):
+                system_prompt_found = True
+                # The system prompt from frontend already has the template, just replace placeholders
+
+                # Format tools
+                tools_str = ""
+                tool_names = []
+                if include_tools and self.tools:
+                    # Convert tools to prompt message tools format
+                    prompt_tools = [tool.to_prompt_message_tool() for tool in self.tools]
+                    tool_names = [tool.name for tool in prompt_tools]
+
+                    # Format tools as JSON for comprehensive information
+                    from core.model_runtime.utils.encoders import jsonable_encoder
+
+                    tools_str = json.dumps(jsonable_encoder(prompt_tools), indent=2)
+                    tool_names_str = ", ".join(f'"{name}"' for name in tool_names)
+                else:
+                    tools_str = "No tools available"
+                    tool_names_str = ""
+
+                # Replace placeholders in the existing system prompt
+                updated_content = msg.content
+                assert isinstance(updated_content, str)
+                updated_content = updated_content.replace("{{instruction}}", instruction)
+                updated_content = updated_content.replace("{{tools}}", tools_str)
+                updated_content = updated_content.replace("{{tool_names}}", tool_names_str)
+
+                # Create new SystemPromptMessage with updated content
+                messages[i] = SystemPromptMessage(content=updated_content)
+                break
+
+        # If no system prompt found, that's unexpected but add scratchpad anyway
+        if not system_prompt_found:
+            # This shouldn't happen if frontend is working correctly
+            pass
+
+        # Format agent scratchpad
+        scratchpad_str = ""
+        if agent_scratchpad:
+            scratchpad_parts: list[str] = []
+            for unit in agent_scratchpad:
+                if unit.thought:
+                    scratchpad_parts.append(f"Thought: {unit.thought}")
+                if unit.action_str:
+                    scratchpad_parts.append(f"Action:\n```\n{unit.action_str}\n```")
+                if unit.observation:
+                    scratchpad_parts.append(f"Observation: {unit.observation}")
+            scratchpad_str = "\n".join(scratchpad_parts)
+
+        # If there's a scratchpad, append it to the last message
+        if scratchpad_str:
+            messages.append(AssistantPromptMessage(content=scratchpad_str))
+
+        return messages
+
+    def _handle_chunks(
+        self,
+        chunks: Union[Generator[LLMResultChunk, None, None], LLMResult],
+        llm_usage: dict[str, Any],
+        model_log: AgentLog,
+        current_messages: list[PromptMessage],
+    ) -> Generator[
+        LLMResultChunk | AgentLog,
+        None,
+        tuple[AgentScratchpadUnit, str | None],
+    ]:
+        """Handle LLM response chunks and extract action/thought.
+
+        Returns a tuple of (scratchpad_unit, finish_reason).
+        """
+        usage_dict: dict[str, Any] = {}
+
+        # Convert non-streaming to streaming format if needed
+        if isinstance(chunks, LLMResult):
+            # Create a generator from the LLMResult
+            def result_to_chunks() -> Generator[LLMResultChunk, None, None]:
+                yield LLMResultChunk(
+                    model=chunks.model,
+                    prompt_messages=chunks.prompt_messages,
+                    delta=LLMResultChunkDelta(
+                        index=0,
+                        message=chunks.message,
+                        usage=chunks.usage,
+                        finish_reason=None,  # LLMResult doesn't have finish_reason, only streaming chunks do
+                    ),
+                    system_fingerprint=chunks.system_fingerprint or "",
+                )
+
+            streaming_chunks = result_to_chunks()
+        else:
+            streaming_chunks = chunks
+
+        react_chunks = CotAgentOutputParser.handle_react_stream_output(streaming_chunks, usage_dict)
+
+        # Initialize scratchpad unit
+        scratchpad = AgentScratchpadUnit(
+            agent_response="",
+            thought="",
+            action_str="",
+            observation="",
+            action=None,
+        )
+
+        finish_reason: str | None = None
+
+        # Process chunks
+        for chunk in react_chunks:
+            if isinstance(chunk, AgentScratchpadUnit.Action):
+                # Action detected
+                action_str = json.dumps(chunk.model_dump())
+                scratchpad.agent_response = (scratchpad.agent_response or "") + action_str
+                scratchpad.action_str = action_str
+                scratchpad.action = chunk
+
+                yield self._create_text_chunk(json.dumps(chunk.model_dump()), current_messages)
+            else:
+                # Text chunk
+                chunk_text = str(chunk)
+                scratchpad.agent_response = (scratchpad.agent_response or "") + chunk_text
+                scratchpad.thought = (scratchpad.thought or "") + chunk_text
+
+                yield self._create_text_chunk(chunk_text, current_messages)
+
+        # Update usage
+        if usage_dict.get("usage"):
+            if llm_usage.get("usage"):
+                self._accumulate_usage(llm_usage, usage_dict["usage"])
+            else:
+                llm_usage["usage"] = usage_dict["usage"]
+
+        # Clean up thought
+        scratchpad.thought = (scratchpad.thought or "").strip() or "I am thinking about how to help you"
+
+        # Finish model log
+        yield self._finish_log(
+            model_log,
+            data={
+                "thought": scratchpad.thought,
+                "action": scratchpad.action_str if scratchpad.action else None,
+            },
+            usage=llm_usage.get("usage"),
+        )
+
+        return scratchpad, finish_reason
+
+    def _handle_tool_call(
+        self,
+        action: AgentScratchpadUnit.Action,
+        prompt_messages: list[PromptMessage],
+        round_log: AgentLog,
+    ) -> Generator[AgentLog, None, tuple[str, list[File]]]:
+        """Handle tool call and return observation with files."""
+        tool_name = action.action_name
+        tool_args: dict[str, Any] | str = action.action_input
+
+        # Find tool instance first to get metadata
+        tool_instance = self._find_tool_by_name(tool_name)
+        tool_metadata = self._get_tool_metadata(tool_instance) if tool_instance else {}
+
+        # Start tool log with tool metadata
+        tool_log = self._create_log(
+            label=f"CALL {tool_name}",
+            log_type=AgentLog.LogType.TOOL_CALL,
+            status=AgentLog.LogStatus.START,
+            data={
+                "tool_name": tool_name,
+                "tool_args": tool_args,
+            },
+            parent_id=round_log.id,
+            extra_metadata=tool_metadata,
+        )
+        yield tool_log
+
+        if not tool_instance:
+            # Finish tool log with error
+            yield self._finish_log(
+                tool_log,
+                data={
+                    **tool_log.data,
+                    "error": f"Tool {tool_name} not found",
+                },
+            )
+            return f"Tool {tool_name} not found", []
+
+        # Ensure tool_args is a dict
+        tool_args_dict: dict[str, Any]
+        if isinstance(tool_args, str):
+            try:
+                tool_args_dict = json.loads(tool_args)
+            except json.JSONDecodeError:
+                tool_args_dict = {"input": tool_args}
+        elif not isinstance(tool_args, dict):
+            tool_args_dict = {"input": str(tool_args)}
+        else:
+            tool_args_dict = tool_args
+
+        # Invoke tool using base class method with error handling
+        try:
+            response_content, tool_files, tool_invoke_meta = self._invoke_tool(tool_instance, tool_args_dict, tool_name)
+
+            # Finish tool log
+            yield self._finish_log(
+                tool_log,
+                data={
+                    **tool_log.data,
+                    "output": response_content,
+                    "files": len(tool_files),
+                    "meta": tool_invoke_meta.to_dict() if tool_invoke_meta else None,
+                },
+            )
+
+            return response_content or "Tool executed successfully", tool_files
+        except Exception as e:
+            # Tool invocation failed, yield error log
+            error_message = str(e)
+            tool_log.status = AgentLog.LogStatus.ERROR
+            tool_log.error = error_message
+            tool_log.data = {
+                **tool_log.data,
+                "error": error_message,
+            }
+            yield tool_log
+
+            return f"Tool execution failed: {error_message}", []

api/core/agent/patterns/strategy_factory.py

@@ -0,0 +1,107 @@
+"""Strategy factory for creating agent strategies."""
+
+from __future__ import annotations
+
+from typing import TYPE_CHECKING
+
+from core.agent.entities import AgentEntity, ExecutionContext
+from core.file.models import File
+from core.model_manager import ModelInstance
+from core.model_runtime.entities.model_entities import ModelFeature
+
+from .base import AgentPattern, ToolInvokeHook
+from .function_call import FunctionCallStrategy
+from .react import ReActStrategy
+
+if TYPE_CHECKING:
+    from core.tools.__base.tool import Tool
+
+
+class StrategyFactory:
+    """Factory for creating agent strategies based on model features."""
+
+    # Tool calling related features
+    TOOL_CALL_FEATURES = {ModelFeature.TOOL_CALL, ModelFeature.MULTI_TOOL_CALL, ModelFeature.STREAM_TOOL_CALL}
+
+    @staticmethod
+    def create_strategy(
+        model_features: list[ModelFeature],
+        model_instance: ModelInstance,
+        context: ExecutionContext,
+        tools: list[Tool],
+        files: list[File],
+        max_iterations: int = 10,
+        workflow_call_depth: int = 0,
+        agent_strategy: AgentEntity.Strategy | None = None,
+        tool_invoke_hook: ToolInvokeHook | None = None,
+        instruction: str = "",
+    ) -> AgentPattern:
+        """
+        Create an appropriate strategy based on model features.
+
+        Args:
+            model_features: List of model features/capabilities
+            model_instance: Model instance to use
+            context: Execution context containing trace/audit information
+            tools: Available tools
+            files: Available files
+            max_iterations: Maximum iterations for the strategy
+            workflow_call_depth: Depth of workflow calls
+            agent_strategy: Optional explicit strategy override
+            tool_invoke_hook: Optional hook for custom tool invocation (e.g., agent_invoke)
+            instruction: Optional instruction for ReAct strategy
+
+        Returns:
+            AgentStrategy instance
+        """
+        # If explicit strategy is provided and it's Function Calling, try to use it if supported
+        if agent_strategy == AgentEntity.Strategy.FUNCTION_CALLING:
+            if set(model_features) & StrategyFactory.TOOL_CALL_FEATURES:
+                return FunctionCallStrategy(
+                    model_instance=model_instance,
+                    context=context,
+                    tools=tools,
+                    files=files,
+                    max_iterations=max_iterations,
+                    workflow_call_depth=workflow_call_depth,
+                    tool_invoke_hook=tool_invoke_hook,
+                )
+            # Fallback to ReAct if FC is requested but not supported
+
+        # If explicit strategy is Chain of Thought (ReAct)
+        if agent_strategy == AgentEntity.Strategy.CHAIN_OF_THOUGHT:
+            return ReActStrategy(
+                model_instance=model_instance,
+                context=context,
+                tools=tools,
+                files=files,
+                max_iterations=max_iterations,
+                workflow_call_depth=workflow_call_depth,
+                tool_invoke_hook=tool_invoke_hook,
+                instruction=instruction,
+            )
+
+        # Default auto-selection logic
+        if set(model_features) & StrategyFactory.TOOL_CALL_FEATURES:
+            # Model supports native function calling
+            return FunctionCallStrategy(
+                model_instance=model_instance,
+                context=context,
+                tools=tools,
+                files=files,
+                max_iterations=max_iterations,
+                workflow_call_depth=workflow_call_depth,
+                tool_invoke_hook=tool_invoke_hook,
+            )
+        else:
+            # Use ReAct strategy for models without function calling
+            return ReActStrategy(
+                model_instance=model_instance,
+                context=context,
+                tools=tools,
+                files=files,
+                max_iterations=max_iterations,
+                workflow_call_depth=workflow_call_depth,
+                tool_invoke_hook=tool_invoke_hook,
+                instruction=instruction,
+            )

api/core/app/apps/advanced_chat/app_generator.py

@@ -24,11 +24,13 @@ from core.app.apps.message_based_app_generator import MessageBasedAppGenerator
 from core.app.apps.message_based_app_queue_manager import MessageBasedAppQueueManager
 from core.app.entities.app_invoke_entities import AdvancedChatAppGenerateEntity, InvokeFrom
 from core.app.entities.task_entities import ChatbotAppBlockingResponse, ChatbotAppStreamResponse
+from core.app.layers.sandbox_layer import SandboxLayer
 from core.helper.trace_id_helper import extract_external_trace_id_from_args
 from core.model_runtime.errors.invoke import InvokeAuthorizationError
 from core.ops.ops_trace_manager import TraceQueueManager
 from core.prompt.utils.get_thread_messages_length import get_thread_messages_length
 from core.repositories import DifyCoreRepositoryFactory
+from core.sandbox import Sandbox, SandboxManager
 from core.workflow.repositories.draft_variable_repository import (
     DraftVariableSaverFactory,
 )
@@ -40,7 +42,9 @@ from factories import file_factory
 from libs.flask_utils import preserve_flask_contexts
 from models import Account, App, Conversation, EndUser, Message, Workflow, WorkflowNodeExecutionTriggeredFrom
 from models.enums import WorkflowRunTriggeredFrom
+from models.workflow_features import WorkflowFeatures
 from services.conversation_service import ConversationService
+from services.sandbox.sandbox_provider_service import SandboxProviderService
 from services.workflow_draft_variable_service import (
     DraftVarLoader,
     WorkflowDraftVariableService,
@@ -512,6 +516,31 @@ class AdvancedChatAppGenerator(MessageBasedAppGenerator):
                 if workflow is None:
                     raise ValueError("Workflow not found")
 
+                sandbox: Sandbox | None = None
+                graph_engine_layers: tuple = ()
+                if workflow.get_feature(WorkflowFeatures.SANDBOX).enabled:
+                    sandbox_provider = SandboxProviderService.get_sandbox_provider(
+                        application_generate_entity.app_config.tenant_id
+                    )
+                    if workflow.version == Workflow.VERSION_DRAFT:
+                        sandbox = SandboxManager.create_draft(
+                            tenant_id=application_generate_entity.app_config.tenant_id,
+                            app_id=application_generate_entity.app_config.app_id,
+                            user_id=application_generate_entity.user_id,
+                            sandbox_provider=sandbox_provider,
+                        )
+                    else:
+                        if application_generate_entity.workflow_run_id is None:
+                            raise ValueError("workflow_run_id is required when sandbox is enabled")
+                        sandbox = SandboxManager.create(
+                            tenant_id=application_generate_entity.app_config.tenant_id,
+                            app_id=application_generate_entity.app_config.app_id,
+                            user_id=application_generate_entity.user_id,
+                            workflow_execution_id=application_generate_entity.workflow_run_id,
+                            sandbox_provider=sandbox_provider,
+                        )
+                    graph_engine_layers = (SandboxLayer(sandbox=sandbox),)
+
                 # Determine system_user_id based on invocation source
                 is_external_api_call = application_generate_entity.invoke_from in {
                     InvokeFrom.WEB_APP,
@@ -542,6 +571,8 @@ class AdvancedChatAppGenerator(MessageBasedAppGenerator):
                 app=app,
                 workflow_execution_repository=workflow_execution_repository,
                 workflow_node_execution_repository=workflow_node_execution_repository,
+                graph_engine_layers=graph_engine_layers,
+                sandbox=sandbox,
             )
 
             try:

api/core/app/apps/advanced_chat/app_runner.py

@@ -24,6 +24,7 @@ from core.app.layers.conversation_variable_persist_layer import ConversationVari
 from core.db.session_factory import session_factory
 from core.moderation.base import ModerationError
 from core.moderation.input_moderation import InputModeration
+from core.sandbox import Sandbox
 from core.variables.variables import Variable
 from core.workflow.enums import WorkflowType
 from core.workflow.graph_engine.command_channels.redis_channel import RedisChannel
@@ -66,6 +67,7 @@ class AdvancedChatAppRunner(WorkflowBasedAppRunner):
         workflow_execution_repository: WorkflowExecutionRepository,
         workflow_node_execution_repository: WorkflowNodeExecutionRepository,
         graph_engine_layers: Sequence[GraphEngineLayer] = (),
+        sandbox: Sandbox | None = None,
     ):
         super().__init__(
             queue_manager=queue_manager,
@@ -82,6 +84,7 @@ class AdvancedChatAppRunner(WorkflowBasedAppRunner):
         self._app = app
         self._workflow_execution_repository = workflow_execution_repository
         self._workflow_node_execution_repository = workflow_node_execution_repository
+        self._sandbox = sandbox
 
     @trace_span(WorkflowAppRunnerHandler)
     def run(self):
@@ -156,6 +159,10 @@ class AdvancedChatAppRunner(WorkflowBasedAppRunner):
 
             # init graph
             graph_runtime_state = GraphRuntimeState(variable_pool=variable_pool, start_at=time.time())
+
+            if self._sandbox:
+                graph_runtime_state.set_sandbox(self._sandbox)
+
             graph = self._init_graph(
                 graph_config=self._workflow.graph_dict,
                 graph_runtime_state=graph_runtime_state,

api/core/app/apps/advanced_chat/generate_response_converter.py

@@ -82,7 +82,7 @@ class AdvancedChatAppGenerateResponseConverter(AppGenerateResponseConverter):
                 data = cls._error_to_stream_response(sub_stream_response.err)
                 response_chunk.update(data)
             else:
-                response_chunk.update(sub_stream_response.model_dump(mode="json"))
+                response_chunk.update(sub_stream_response.model_dump(mode="json", exclude_none=True))
             yield response_chunk
 
     @classmethod
@@ -110,7 +110,7 @@ class AdvancedChatAppGenerateResponseConverter(AppGenerateResponseConverter):
             }
 
             if isinstance(sub_stream_response, MessageEndStreamResponse):
-                sub_stream_response_dict = sub_stream_response.model_dump(mode="json")
+                sub_stream_response_dict = sub_stream_response.model_dump(mode="json", exclude_none=True)
                 metadata = sub_stream_response_dict.get("metadata", {})
                 sub_stream_response_dict["metadata"] = cls._get_simple_metadata(metadata)
                 response_chunk.update(sub_stream_response_dict)
@@ -120,6 +120,6 @@ class AdvancedChatAppGenerateResponseConverter(AppGenerateResponseConverter):
             elif isinstance(sub_stream_response, NodeStartStreamResponse | NodeFinishStreamResponse):
                 response_chunk.update(sub_stream_response.to_ignore_detail_dict())
             else:
-                response_chunk.update(sub_stream_response.model_dump(mode="json"))
+                response_chunk.update(sub_stream_response.model_dump(mode="json", exclude_none=True))
 
             yield response_chunk

api/core/app/apps/advanced_chat/generate_task_pipeline.py

@@ -4,6 +4,7 @@ import re
 import time
 from collections.abc import Callable, Generator, Mapping
 from contextlib import contextmanager
+from dataclasses import dataclass, field
 from threading import Thread
 from typing import Any, Union
 
@@ -19,6 +20,7 @@ from core.app.entities.app_invoke_entities import (
     InvokeFrom,
 )
 from core.app.entities.queue_entities import (
+    ChunkType,
     MessageQueueMessage,
     QueueAdvancedChatMessageEndEvent,
     QueueAgentLogEvent,
@@ -70,13 +72,131 @@ from core.workflow.runtime import GraphRuntimeState
 from core.workflow.system_variable import SystemVariable
 from extensions.ext_database import db
 from libs.datetime_utils import naive_utc_now
-from models import Account, Conversation, EndUser, Message, MessageFile
+from models import Account, Conversation, EndUser, LLMGenerationDetail, Message, MessageFile
 from models.enums import CreatorUserRole
 from models.workflow import Workflow
 
 logger = logging.getLogger(__name__)
 
 
+@dataclass
+class StreamEventBuffer:
+    """
+    Buffer for recording stream events in order to reconstruct the generation sequence.
+    Records the exact order of text chunks, thoughts, and tool calls as they stream.
+    """
+
+    # Accumulated reasoning content (each thought block is a separate element)
+    reasoning_content: list[str] = field(default_factory=list)
+    # Current reasoning buffer (accumulates until we see a different event type)
+    _current_reasoning: str = ""
+    # Tool calls with their details
+    tool_calls: list[dict] = field(default_factory=list)
+    # Tool call ID to index mapping for updating results
+    _tool_call_id_map: dict[str, int] = field(default_factory=dict)
+    # Sequence of events in stream order
+    sequence: list[dict] = field(default_factory=list)
+    # Current position in answer text
+    _content_position: int = 0
+    # Track last event type to detect transitions
+    _last_event_type: str | None = None
+
+    def _flush_current_reasoning(self) -> None:
+        """Flush accumulated reasoning to the list and add to sequence."""
+        if self._current_reasoning.strip():
+            self.reasoning_content.append(self._current_reasoning.strip())
+            self.sequence.append({"type": "reasoning", "index": len(self.reasoning_content) - 1})
+            self._current_reasoning = ""
+
+    def record_text_chunk(self, text: str) -> None:
+        """Record a text chunk event."""
+        if not text:
+            return
+
+        # Flush any pending reasoning first
+        if self._last_event_type == "thought":
+            self._flush_current_reasoning()
+
+        text_len = len(text)
+        start_pos = self._content_position
+
+        # If last event was also content, extend it; otherwise create new
+        if self.sequence and self.sequence[-1].get("type") == "content":
+            self.sequence[-1]["end"] = start_pos + text_len
+        else:
+            self.sequence.append({"type": "content", "start": start_pos, "end": start_pos + text_len})
+
+        self._content_position += text_len
+        self._last_event_type = "content"
+
+    def record_thought_chunk(self, text: str) -> None:
+        """Record a thought/reasoning chunk event."""
+        if not text:
+            return
+
+        # Accumulate thought content
+        self._current_reasoning += text
+        self._last_event_type = "thought"
+
+    def record_tool_call(
+        self,
+        tool_call_id: str,
+        tool_name: str,
+        tool_arguments: str,
+        tool_icon: str | dict | None = None,
+        tool_icon_dark: str | dict | None = None,
+    ) -> None:
+        """Record a tool call event."""
+        if not tool_call_id:
+            return
+
+        # Flush any pending reasoning first
+        if self._last_event_type == "thought":
+            self._flush_current_reasoning()
+
+        # Check if this tool call already exists (we might get multiple chunks)
+        if tool_call_id in self._tool_call_id_map:
+            idx = self._tool_call_id_map[tool_call_id]
+            # Update arguments if provided
+            if tool_arguments:
+                self.tool_calls[idx]["arguments"] = tool_arguments
+        else:
+            # New tool call
+            tool_call = {
+                "id": tool_call_id or "",
+                "name": tool_name or "",
+                "arguments": tool_arguments or "",
+                "result": "",
+                "elapsed_time": None,
+                "icon": tool_icon,
+                "icon_dark": tool_icon_dark,
+            }
+            self.tool_calls.append(tool_call)
+            idx = len(self.tool_calls) - 1
+            self._tool_call_id_map[tool_call_id] = idx
+            self.sequence.append({"type": "tool_call", "index": idx})
+
+        self._last_event_type = "tool_call"
+
+    def record_tool_result(self, tool_call_id: str, result: str, tool_elapsed_time: float | None = None) -> None:
+        """Record a tool result event (update existing tool call)."""
+        if not tool_call_id:
+            return
+        if tool_call_id in self._tool_call_id_map:
+            idx = self._tool_call_id_map[tool_call_id]
+            self.tool_calls[idx]["result"] = result
+            self.tool_calls[idx]["elapsed_time"] = tool_elapsed_time
+
+    def finalize(self) -> None:
+        """Finalize the buffer, flushing any pending data."""
+        if self._last_event_type == "thought":
+            self._flush_current_reasoning()
+
+    def has_data(self) -> bool:
+        """Check if there's any meaningful data recorded."""
+        return bool(self.reasoning_content or self.tool_calls or self.sequence)
+
+
 class AdvancedChatAppGenerateTaskPipeline(GraphRuntimeStateSupport):
     """
     AdvancedChatAppGenerateTaskPipeline is a class that generate stream output and state management for Application.
@@ -144,6 +264,8 @@ class AdvancedChatAppGenerateTaskPipeline(GraphRuntimeStateSupport):
         self._workflow_run_id: str = ""
         self._draft_var_saver_factory = draft_var_saver_factory
         self._graph_runtime_state: GraphRuntimeState | None = None
+        # Stream event buffer for recording generation sequence
+        self._stream_buffer = StreamEventBuffer()
         self._seed_graph_runtime_state_from_queue_manager()
 
     def process(self) -> Union[ChatbotAppBlockingResponse, Generator[ChatbotAppStreamResponse, None, None]]:
@@ -383,7 +505,7 @@ class AdvancedChatAppGenerateTaskPipeline(GraphRuntimeStateSupport):
         queue_message: Union[WorkflowQueueMessage, MessageQueueMessage] | None = None,
         **kwargs,
     ) -> Generator[StreamResponse, None, None]:
-        """Handle text chunk events."""
+        """Handle text chunk events and record to stream buffer for sequence reconstruction."""
         delta_text = event.text
         if delta_text is None:
             return
@@ -405,9 +527,53 @@ class AdvancedChatAppGenerateTaskPipeline(GraphRuntimeStateSupport):
         if tts_publisher and queue_message:
             tts_publisher.publish(queue_message)
 
-        self._task_state.answer += delta_text
+        tool_call = event.tool_call
+        tool_result = event.tool_result
+        tool_payload = tool_call or tool_result
+        tool_call_id = tool_payload.id if tool_payload and tool_payload.id else ""
+        tool_name = tool_payload.name if tool_payload and tool_payload.name else ""
+        tool_arguments = tool_call.arguments if tool_call and tool_call.arguments else ""
+        tool_files = tool_result.files if tool_result else []
+        tool_elapsed_time = tool_result.elapsed_time if tool_result else None
+        tool_icon = tool_payload.icon if tool_payload else None
+        tool_icon_dark = tool_payload.icon_dark if tool_payload else None
+        # Record stream event based on chunk type
+        chunk_type = event.chunk_type or ChunkType.TEXT
+        match chunk_type:
+            case ChunkType.TEXT:
+                self._stream_buffer.record_text_chunk(delta_text)
+                self._task_state.answer += delta_text
+            case ChunkType.THOUGHT:
+                # Reasoning should not be part of final answer text
+                self._stream_buffer.record_thought_chunk(delta_text)
+            case ChunkType.TOOL_CALL:
+                self._stream_buffer.record_tool_call(
+                    tool_call_id=tool_call_id,
+                    tool_name=tool_name,
+                    tool_arguments=tool_arguments,
+                    tool_icon=tool_icon,
+                    tool_icon_dark=tool_icon_dark,
+                )
+            case ChunkType.TOOL_RESULT:
+                self._stream_buffer.record_tool_result(
+                    tool_call_id=tool_call_id,
+                    result=delta_text,
+                    tool_elapsed_time=tool_elapsed_time,
+                )
+            case _:
+                pass
         yield self._message_cycle_manager.message_to_stream_response(
-            answer=delta_text, message_id=self._message_id, from_variable_selector=event.from_variable_selector
+            answer=delta_text,
+            message_id=self._message_id,
+            from_variable_selector=event.from_variable_selector,
+            chunk_type=event.chunk_type.value if event.chunk_type else None,
+            tool_call_id=tool_call_id or None,
+            tool_name=tool_name or None,
+            tool_arguments=tool_arguments or None,
+            tool_files=tool_files,
+            tool_elapsed_time=tool_elapsed_time,
+            tool_icon=tool_icon,
+            tool_icon_dark=tool_icon_dark,
         )
 
     def _handle_iteration_start_event(
@@ -775,6 +941,7 @@ class AdvancedChatAppGenerateTaskPipeline(GraphRuntimeStateSupport):
 
         # If there are assistant files, remove markdown image links from answer
         answer_text = self._task_state.answer
+        answer_text = self._strip_think_blocks(answer_text)
         if self._recorded_files:
             # Remove markdown image links since we're storing files separately
             answer_text = re.sub(r"!\[.*?\]\(.*?\)", "", answer_text).strip()
@@ -826,6 +993,54 @@ class AdvancedChatAppGenerateTaskPipeline(GraphRuntimeStateSupport):
         ]
         session.add_all(message_files)
 
+        # Save generation detail (reasoning/tool calls/sequence) from stream buffer
+        self._save_generation_detail(session=session, message=message)
+
+    @staticmethod
+    def _strip_think_blocks(text: str) -> str:
+        """Remove <think>...</think> blocks (including their content) from text."""
+        if not text or "<think" not in text.lower():
+            return text
+
+        clean_text = re.sub(r"<think[^>]*>.*?</think>", "", text, flags=re.IGNORECASE | re.DOTALL)
+        clean_text = re.sub(r"\n\s*\n", "\n\n", clean_text).strip()
+        return clean_text
+
+    def _save_generation_detail(self, *, session: Session, message: Message) -> None:
+        """
+        Save LLM generation detail for Chatflow using stream event buffer.
+        The buffer records the exact order of events as they streamed,
+        allowing accurate reconstruction of the generation sequence.
+        """
+        # Finalize the stream buffer to flush any pending data
+        self._stream_buffer.finalize()
+
+        # Only save if there's meaningful data
+        if not self._stream_buffer.has_data():
+            return
+
+        reasoning_content = self._stream_buffer.reasoning_content
+        tool_calls = self._stream_buffer.tool_calls
+        sequence = self._stream_buffer.sequence
+
+        # Check if generation detail already exists for this message
+        existing = session.query(LLMGenerationDetail).filter_by(message_id=message.id).first()
+
+        if existing:
+            existing.reasoning_content = json.dumps(reasoning_content) if reasoning_content else None
+            existing.tool_calls = json.dumps(tool_calls) if tool_calls else None
+            existing.sequence = json.dumps(sequence) if sequence else None
+        else:
+            generation_detail = LLMGenerationDetail(
+                tenant_id=self._application_generate_entity.app_config.tenant_id,
+                app_id=self._application_generate_entity.app_config.app_id,
+                message_id=message.id,
+                reasoning_content=json.dumps(reasoning_content) if reasoning_content else None,
+                tool_calls=json.dumps(tool_calls) if tool_calls else None,
+                sequence=json.dumps(sequence) if sequence else None,
+            )
+            session.add(generation_detail)
+
     def _seed_graph_runtime_state_from_queue_manager(self) -> None:
         """Bootstrap the cached runtime state from the queue manager when present."""
         candidate = self._base_task_pipeline.queue_manager.graph_runtime_state

api/core/app/apps/agent_chat/app_runner.py

@@ -3,10 +3,8 @@ from typing import cast
 
 from sqlalchemy import select
 
-from core.agent.cot_chat_agent_runner import CotChatAgentRunner
-from core.agent.cot_completion_agent_runner import CotCompletionAgentRunner
+from core.agent.agent_app_runner import AgentAppRunner
 from core.agent.entities import AgentEntity
-from core.agent.fc_agent_runner import FunctionCallAgentRunner
 from core.app.apps.agent_chat.app_config_manager import AgentChatAppConfig
 from core.app.apps.base_app_queue_manager import AppQueueManager, PublishFrom
 from core.app.apps.base_app_runner import AppRunner
@@ -14,8 +12,7 @@ from core.app.entities.app_invoke_entities import AgentChatAppGenerateEntity
 from core.app.entities.queue_entities import QueueAnnotationReplyEvent
 from core.memory.token_buffer_memory import TokenBufferMemory
 from core.model_manager import ModelInstance
-from core.model_runtime.entities.llm_entities import LLMMode
-from core.model_runtime.entities.model_entities import ModelFeature, ModelPropertyKey
+from core.model_runtime.entities.model_entities import ModelFeature
 from core.model_runtime.model_providers.__base.large_language_model import LargeLanguageModel
 from core.moderation.base import ModerationError
 from extensions.ext_database import db
@@ -194,22 +191,7 @@ class AgentChatAppRunner(AppRunner):
             raise ValueError("Message not found")
         db.session.close()
 
-        runner_cls: type[FunctionCallAgentRunner] | type[CotChatAgentRunner] | type[CotCompletionAgentRunner]
-        # start agent runner
-        if agent_entity.strategy == AgentEntity.Strategy.CHAIN_OF_THOUGHT:
-            # check LLM mode
-            if model_schema.model_properties.get(ModelPropertyKey.MODE) == LLMMode.CHAT:
-                runner_cls = CotChatAgentRunner
-            elif model_schema.model_properties.get(ModelPropertyKey.MODE) == LLMMode.COMPLETION:
-                runner_cls = CotCompletionAgentRunner
-            else:
-                raise ValueError(f"Invalid LLM mode: {model_schema.model_properties.get(ModelPropertyKey.MODE)}")
-        elif agent_entity.strategy == AgentEntity.Strategy.FUNCTION_CALLING:
-            runner_cls = FunctionCallAgentRunner
-        else:
-            raise ValueError(f"Invalid agent strategy: {agent_entity.strategy}")
-
-        runner = runner_cls(
+        runner = AgentAppRunner(
             tenant_id=app_config.tenant_id,
             application_generate_entity=application_generate_entity,
             conversation=conversation_result,

api/core/app/apps/agent_chat/generate_response_converter.py

@@ -81,7 +81,7 @@ class AgentChatAppGenerateResponseConverter(AppGenerateResponseConverter):
                 data = cls._error_to_stream_response(sub_stream_response.err)
                 response_chunk.update(data)
             else:
-                response_chunk.update(sub_stream_response.model_dump(mode="json"))
+                response_chunk.update(sub_stream_response.model_dump(mode="json", exclude_none=True))
             yield response_chunk
 
     @classmethod
@@ -109,7 +109,7 @@ class AgentChatAppGenerateResponseConverter(AppGenerateResponseConverter):
             }
 
             if isinstance(sub_stream_response, MessageEndStreamResponse):
-                sub_stream_response_dict = sub_stream_response.model_dump(mode="json")
+                sub_stream_response_dict = sub_stream_response.model_dump(mode="json", exclude_none=True)
                 metadata = sub_stream_response_dict.get("metadata", {})
                 sub_stream_response_dict["metadata"] = cls._get_simple_metadata(metadata)
                 response_chunk.update(sub_stream_response_dict)
@@ -117,6 +117,6 @@ class AgentChatAppGenerateResponseConverter(AppGenerateResponseConverter):
                 data = cls._error_to_stream_response(sub_stream_response.err)
                 response_chunk.update(data)
             else:
-                response_chunk.update(sub_stream_response.model_dump(mode="json"))
+                response_chunk.update(sub_stream_response.model_dump(mode="json", exclude_none=True))
 
             yield response_chunk

api/core/app/apps/chat/generate_response_converter.py

@@ -81,7 +81,7 @@ class ChatAppGenerateResponseConverter(AppGenerateResponseConverter):
                 data = cls._error_to_stream_response(sub_stream_response.err)
                 response_chunk.update(data)
             else:
-                response_chunk.update(sub_stream_response.model_dump(mode="json"))
+                response_chunk.update(sub_stream_response.model_dump(mode="json", exclude_none=True))
             yield response_chunk
 
     @classmethod
@@ -109,7 +109,7 @@ class ChatAppGenerateResponseConverter(AppGenerateResponseConverter):
             }
 
             if isinstance(sub_stream_response, MessageEndStreamResponse):
-                sub_stream_response_dict = sub_stream_response.model_dump(mode="json")
+                sub_stream_response_dict = sub_stream_response.model_dump(mode="json", exclude_none=True)
                 metadata = sub_stream_response_dict.get("metadata", {})
                 sub_stream_response_dict["metadata"] = cls._get_simple_metadata(metadata)
                 response_chunk.update(sub_stream_response_dict)
@@ -117,6 +117,6 @@ class ChatAppGenerateResponseConverter(AppGenerateResponseConverter):
                 data = cls._error_to_stream_response(sub_stream_response.err)
                 response_chunk.update(data)
             else:
-                response_chunk.update(sub_stream_response.model_dump(mode="json"))
+                response_chunk.update(sub_stream_response.model_dump(mode="json", exclude_none=True))
 
             yield response_chunk

api/core/app/apps/common/workflow_response_converter.py

@@ -70,6 +70,8 @@ class _NodeSnapshot:
     """Empty string means the node is not executing inside an iteration."""
     loop_id: str = ""
     """Empty string means the node is not executing inside a loop."""
+    parent_node_id: str = ""
+    """Empty string means the node is not an nested node (extractor node)."""
 
 
 class WorkflowResponseConverter:
@@ -131,6 +133,7 @@ class WorkflowResponseConverter:
             start_at=event.start_at,
             iteration_id=event.in_iteration_id or "",
             loop_id=event.in_loop_id or "",
+            parent_node_id=event.in_parent_node_id or "",
         )
         node_execution_id = NodeExecutionId(event.node_execution_id)
         self._node_snapshots[node_execution_id] = snapshot
@@ -287,6 +290,7 @@ class WorkflowResponseConverter:
                 created_at=int(snapshot.start_at.timestamp()),
                 iteration_id=event.in_iteration_id,
                 loop_id=event.in_loop_id,
+                parent_node_id=event.in_parent_node_id,
                 agent_strategy=event.agent_strategy,
             ),
         )
@@ -373,6 +377,7 @@ class WorkflowResponseConverter:
                 files=self.fetch_files_from_node_outputs(event.outputs or {}),
                 iteration_id=event.in_iteration_id,
                 loop_id=event.in_loop_id,
+                parent_node_id=event.in_parent_node_id,
             ),
         )
 
@@ -422,6 +427,7 @@ class WorkflowResponseConverter:
                 files=self.fetch_files_from_node_outputs(event.outputs or {}),
                 iteration_id=event.in_iteration_id,
                 loop_id=event.in_loop_id,
+                parent_node_id=event.in_parent_node_id,
                 retry_index=event.retry_index,
             ),
         )
@@ -671,7 +677,7 @@ class WorkflowResponseConverter:
             task_id=task_id,
             data=AgentLogStreamResponse.Data(
                 node_execution_id=event.node_execution_id,
-                id=event.id,
+                message_id=event.id,
                 parent_id=event.parent_id,
                 label=event.label,
                 error=event.error,

api/core/app/apps/completion/generate_response_converter.py

@@ -79,7 +79,7 @@ class CompletionAppGenerateResponseConverter(AppGenerateResponseConverter):
                 data = cls._error_to_stream_response(sub_stream_response.err)
                 response_chunk.update(data)
             else:
-                response_chunk.update(sub_stream_response.model_dump(mode="json"))
+                response_chunk.update(sub_stream_response.model_dump(mode="json", exclude_none=True))
             yield response_chunk
 
     @classmethod
@@ -106,7 +106,7 @@ class CompletionAppGenerateResponseConverter(AppGenerateResponseConverter):
             }
 
             if isinstance(sub_stream_response, MessageEndStreamResponse):
-                sub_stream_response_dict = sub_stream_response.model_dump(mode="json")
+                sub_stream_response_dict = sub_stream_response.model_dump(mode="json", exclude_none=True)
                 metadata = sub_stream_response_dict.get("metadata", {})
                 if not isinstance(metadata, dict):
                     metadata = {}
@@ -116,6 +116,6 @@ class CompletionAppGenerateResponseConverter(AppGenerateResponseConverter):
                 data = cls._error_to_stream_response(sub_stream_response.err)
                 response_chunk.update(data)
             else:
-                response_chunk.update(sub_stream_response.model_dump(mode="json"))
+                response_chunk.update(sub_stream_response.model_dump(mode="json", exclude_none=True))
 
             yield response_chunk

api/core/app/apps/pipeline/generate_response_converter.py

@@ -60,7 +60,7 @@ class WorkflowAppGenerateResponseConverter(AppGenerateResponseConverter):
                 data = cls._error_to_stream_response(sub_stream_response.err)
                 response_chunk.update(cast(dict, data))
             else:
-                response_chunk.update(sub_stream_response.model_dump())
+                response_chunk.update(sub_stream_response.model_dump(exclude_none=True))
             yield response_chunk
 
     @classmethod
@@ -91,5 +91,5 @@ class WorkflowAppGenerateResponseConverter(AppGenerateResponseConverter):
             elif isinstance(sub_stream_response, NodeStartStreamResponse | NodeFinishStreamResponse):
                 response_chunk.update(cast(dict, sub_stream_response.to_ignore_detail_dict()))
             else:
-                response_chunk.update(sub_stream_response.model_dump())
+                response_chunk.update(sub_stream_response.model_dump(exclude_none=True))
             yield response_chunk

api/core/app/apps/workflow/app_generator.py

@@ -23,11 +23,13 @@ from core.app.apps.workflow.generate_response_converter import WorkflowAppGenera
 from core.app.apps.workflow.generate_task_pipeline import WorkflowAppGenerateTaskPipeline
 from core.app.entities.app_invoke_entities import InvokeFrom, WorkflowAppGenerateEntity
 from core.app.entities.task_entities import WorkflowAppBlockingResponse, WorkflowAppStreamResponse
+from core.app.layers.sandbox_layer import SandboxLayer
 from core.db.session_factory import session_factory
 from core.helper.trace_id_helper import extract_external_trace_id_from_args
 from core.model_runtime.errors.invoke import InvokeAuthorizationError
 from core.ops.ops_trace_manager import TraceQueueManager
 from core.repositories import DifyCoreRepositoryFactory
+from core.sandbox import Sandbox, SandboxManager
 from core.workflow.graph_engine.layers.base import GraphEngineLayer
 from core.workflow.repositories.draft_variable_repository import DraftVariableSaverFactory
 from core.workflow.repositories.workflow_execution_repository import WorkflowExecutionRepository
@@ -38,6 +40,8 @@ from factories import file_factory
 from libs.flask_utils import preserve_flask_contexts
 from models import Account, App, EndUser, Workflow, WorkflowNodeExecutionTriggeredFrom
 from models.enums import WorkflowRunTriggeredFrom
+from models.workflow_features import WorkflowFeatures
+from services.sandbox.sandbox_provider_service import SandboxProviderService
 from services.workflow_draft_variable_service import DraftVarLoader, WorkflowDraftVariableService
 
 SKIP_PREPARE_USER_INPUTS_KEY = "_skip_prepare_user_inputs"
@@ -488,6 +492,31 @@ class WorkflowAppGenerator(BaseAppGenerator):
                 if workflow is None:
                     raise ValueError("Workflow not found")
 
+                sandbox: Sandbox | None = None
+                if workflow.get_feature(WorkflowFeatures.SANDBOX).enabled:
+                    sandbox_provider = SandboxProviderService.get_sandbox_provider(
+                        application_generate_entity.app_config.tenant_id
+                    )
+                    if workflow.version == Workflow.VERSION_DRAFT:
+                        sandbox = SandboxManager.create_draft(
+                            tenant_id=application_generate_entity.app_config.tenant_id,
+                            app_id=application_generate_entity.app_config.app_id,
+                            user_id=application_generate_entity.user_id,
+                            sandbox_provider=sandbox_provider,
+                        )
+                    else:
+                        sandbox = SandboxManager.create(
+                            tenant_id=application_generate_entity.app_config.tenant_id,
+                            app_id=application_generate_entity.app_config.app_id,
+                            user_id=application_generate_entity.user_id,
+                            workflow_execution_id=application_generate_entity.workflow_execution_id,
+                            sandbox_provider=sandbox_provider,
+                        )
+                    graph_engine_layers = (
+                        *graph_engine_layers,
+                        SandboxLayer(sandbox=sandbox),
+                    )
+
                 # Determine system_user_id based on invocation source
                 is_external_api_call = application_generate_entity.invoke_from in {
                     InvokeFrom.WEB_APP,
@@ -512,6 +541,7 @@ class WorkflowAppGenerator(BaseAppGenerator):
                 workflow_node_execution_repository=workflow_node_execution_repository,
                 root_node_id=root_node_id,
                 graph_engine_layers=graph_engine_layers,
+                sandbox=sandbox,
             )
 
             try:

api/core/app/apps/workflow/app_runner.py

@@ -7,6 +7,7 @@ from core.app.apps.base_app_queue_manager import AppQueueManager
 from core.app.apps.workflow.app_config_manager import WorkflowAppConfig
 from core.app.apps.workflow_app_runner import WorkflowBasedAppRunner
 from core.app.entities.app_invoke_entities import InvokeFrom, WorkflowAppGenerateEntity
+from core.sandbox import Sandbox
 from core.workflow.enums import WorkflowType
 from core.workflow.graph_engine.command_channels.redis_channel import RedisChannel
 from core.workflow.graph_engine.layers.base import GraphEngineLayer
@@ -42,6 +43,7 @@ class WorkflowAppRunner(WorkflowBasedAppRunner):
         workflow_execution_repository: WorkflowExecutionRepository,
         workflow_node_execution_repository: WorkflowNodeExecutionRepository,
         graph_engine_layers: Sequence[GraphEngineLayer] = (),
+        sandbox: Sandbox | None = None,
     ):
         super().__init__(
             queue_manager=queue_manager,
@@ -55,6 +57,7 @@ class WorkflowAppRunner(WorkflowBasedAppRunner):
         self._root_node_id = root_node_id
         self._workflow_execution_repository = workflow_execution_repository
         self._workflow_node_execution_repository = workflow_node_execution_repository
+        self._sandbox = sandbox
 
     @trace_span(WorkflowAppRunnerHandler)
     def run(self):
@@ -99,6 +102,9 @@ class WorkflowAppRunner(WorkflowBasedAppRunner):
 
             graph_runtime_state = GraphRuntimeState(variable_pool=variable_pool, start_at=time.perf_counter())
 
+            if self._sandbox:
+                graph_runtime_state.set_sandbox(self._sandbox)
+
             # init graph
             graph = self._init_graph(
                 graph_config=self._workflow.graph_dict,

api/core/app/apps/workflow/generate_response_converter.py

@@ -60,7 +60,7 @@ class WorkflowAppGenerateResponseConverter(AppGenerateResponseConverter):
                 data = cls._error_to_stream_response(sub_stream_response.err)
                 response_chunk.update(data)
             else:
-                response_chunk.update(sub_stream_response.model_dump(mode="json"))
+                response_chunk.update(sub_stream_response.model_dump(mode="json", exclude_none=True))
             yield response_chunk
 
     @classmethod
@@ -91,5 +91,5 @@ class WorkflowAppGenerateResponseConverter(AppGenerateResponseConverter):
             elif isinstance(sub_stream_response, NodeStartStreamResponse | NodeFinishStreamResponse):
                 response_chunk.update(sub_stream_response.to_ignore_detail_dict())
             else:
-                response_chunk.update(sub_stream_response.model_dump(mode="json"))
+                response_chunk.update(sub_stream_response.model_dump(mode="json", exclude_none=True))
             yield response_chunk

api/core/app/apps/workflow/generate_task_pipeline.py

@@ -13,6 +13,7 @@ from core.app.apps.common.workflow_response_converter import WorkflowResponseCon
 from core.app.entities.app_invoke_entities import InvokeFrom, WorkflowAppGenerateEntity
 from core.app.entities.queue_entities import (
     AppQueueEvent,
+    ChunkType,
     MessageQueueMessage,
     QueueAgentLogEvent,
     QueueErrorEvent,
@@ -483,11 +484,33 @@ class WorkflowAppGenerateTaskPipeline(GraphRuntimeStateSupport):
         if delta_text is None:
             return
 
+        tool_call = event.tool_call
+        tool_result = event.tool_result
+        tool_payload = tool_call or tool_result
+        tool_call_id = tool_payload.id if tool_payload and tool_payload.id else None
+        tool_name = tool_payload.name if tool_payload and tool_payload.name else None
+        tool_arguments = tool_call.arguments if tool_call else None
+        tool_elapsed_time = tool_result.elapsed_time if tool_result else None
+        tool_files = tool_result.files if tool_result else []
+        tool_icon = tool_payload.icon if tool_payload else None
+        tool_icon_dark = tool_payload.icon_dark if tool_payload else None
+
         # only publish tts message at text chunk streaming
         if tts_publisher and queue_message:
             tts_publisher.publish(queue_message)
 
-        yield self._text_chunk_to_stream_response(delta_text, from_variable_selector=event.from_variable_selector)
+        yield self._text_chunk_to_stream_response(
+            text=delta_text,
+            from_variable_selector=event.from_variable_selector,
+            chunk_type=event.chunk_type,
+            tool_call_id=tool_call_id,
+            tool_name=tool_name,
+            tool_arguments=tool_arguments,
+            tool_files=tool_files,
+            tool_elapsed_time=tool_elapsed_time,
+            tool_icon=tool_icon,
+            tool_icon_dark=tool_icon_dark,
+        )
 
     def _handle_agent_log_event(self, event: QueueAgentLogEvent, **kwargs) -> Generator[StreamResponse, None, None]:
         """Handle agent log events."""
@@ -650,16 +673,61 @@ class WorkflowAppGenerateTaskPipeline(GraphRuntimeStateSupport):
         session.add(workflow_app_log)
 
     def _text_chunk_to_stream_response(
-        self, text: str, from_variable_selector: list[str] | None = None
+        self,
+        text: str,
+        from_variable_selector: list[str] | None = None,
+        chunk_type: ChunkType | None = None,
+        tool_call_id: str | None = None,
+        tool_name: str | None = None,
+        tool_arguments: str | None = None,
+        tool_files: list[str] | None = None,
+        tool_error: str | None = None,
+        tool_elapsed_time: float | None = None,
+        tool_icon: str | dict | None = None,
+        tool_icon_dark: str | dict | None = None,
     ) -> TextChunkStreamResponse:
         """
         Handle completed event.
         :param text: text
         :return:
         """
+        from core.app.entities.task_entities import ChunkType as ResponseChunkType
+
+        response_chunk_type = ResponseChunkType(chunk_type.value) if chunk_type else ResponseChunkType.TEXT
+
+        data = TextChunkStreamResponse.Data(
+            text=text,
+            from_variable_selector=from_variable_selector,
+            chunk_type=response_chunk_type,
+        )
+
+        if response_chunk_type == ResponseChunkType.TOOL_CALL:
+            data = data.model_copy(
+                update={
+                    "tool_call_id": tool_call_id,
+                    "tool_name": tool_name,
+                    "tool_arguments": tool_arguments,
+                    "tool_icon": tool_icon,
+                    "tool_icon_dark": tool_icon_dark,
+                }
+            )
+        elif response_chunk_type == ResponseChunkType.TOOL_RESULT:
+            data = data.model_copy(
+                update={
+                    "tool_call_id": tool_call_id,
+                    "tool_name": tool_name,
+                    "tool_arguments": tool_arguments,
+                    "tool_files": tool_files,
+                    "tool_error": tool_error,
+                    "tool_elapsed_time": tool_elapsed_time,
+                    "tool_icon": tool_icon,
+                    "tool_icon_dark": tool_icon_dark,
+                }
+            )
+
         response = TextChunkStreamResponse(
             task_id=self._application_generate_entity.task_id,
-            data=TextChunkStreamResponse.Data(text=text, from_variable_selector=from_variable_selector),
+            data=data,
         )
 
         return response

api/core/app/apps/workflow_app_runner.py

@@ -385,6 +385,7 @@ class WorkflowBasedAppRunner:
                     start_at=event.start_at,
                     in_iteration_id=event.in_iteration_id,
                     in_loop_id=event.in_loop_id,
+                    in_parent_node_id=event.in_parent_node_id,
                     inputs=inputs,
                     process_data=process_data,
                     outputs=outputs,
@@ -405,6 +406,7 @@ class WorkflowBasedAppRunner:
                     start_at=event.start_at,
                     in_iteration_id=event.in_iteration_id,
                     in_loop_id=event.in_loop_id,
+                    in_parent_node_id=event.in_parent_node_id,
                     agent_strategy=event.agent_strategy,
                     provider_type=event.provider_type,
                     provider_id=event.provider_id,
@@ -428,6 +430,7 @@ class WorkflowBasedAppRunner:
                     execution_metadata=execution_metadata,
                     in_iteration_id=event.in_iteration_id,
                     in_loop_id=event.in_loop_id,
+                    in_parent_node_id=event.in_parent_node_id,
                 )
             )
         elif isinstance(event, NodeRunFailedEvent):
@@ -444,6 +447,7 @@ class WorkflowBasedAppRunner:
                     execution_metadata=event.node_run_result.metadata,
                     in_iteration_id=event.in_iteration_id,
                     in_loop_id=event.in_loop_id,
+                    in_parent_node_id=event.in_parent_node_id,
                 )
             )
         elif isinstance(event, NodeRunExceptionEvent):
@@ -460,15 +464,25 @@ class WorkflowBasedAppRunner:
                     execution_metadata=event.node_run_result.metadata,
                     in_iteration_id=event.in_iteration_id,
                     in_loop_id=event.in_loop_id,
+                    in_parent_node_id=event.in_parent_node_id,
                 )
             )
         elif isinstance(event, NodeRunStreamChunkEvent):
+            from core.app.entities.queue_entities import ChunkType as QueueChunkType
+
+            if event.is_final and not event.chunk:
+                return
+
             self._publish_event(
                 QueueTextChunkEvent(
                     text=event.chunk,
                     from_variable_selector=list(event.selector),
                     in_iteration_id=event.in_iteration_id,
                     in_loop_id=event.in_loop_id,
+                    chunk_type=QueueChunkType(event.chunk_type.value),
+                    tool_call=event.tool_call,
+                    tool_result=event.tool_result,
+                    in_parent_node_id=event.in_parent_node_id,
                 )
             )
         elif isinstance(event, NodeRunRetrieverResourceEvent):
@@ -477,6 +491,7 @@ class WorkflowBasedAppRunner:
                     retriever_resources=event.retriever_resources,
                     in_iteration_id=event.in_iteration_id,
                     in_loop_id=event.in_loop_id,
+                    in_parent_node_id=event.in_parent_node_id,
                 )
             )
         elif isinstance(event, NodeRunAgentLogEvent):

api/core/app/entities/app_asset_entities.py

@@ -0,0 +1,294 @@
+from __future__ import annotations
+
+from collections import defaultdict
+from collections.abc import Generator
+from enum import StrEnum
+
+from pydantic import BaseModel, Field
+
+
+class AssetNodeType(StrEnum):
+    FILE = "file"
+    FOLDER = "folder"
+
+
+class AppAssetNode(BaseModel):
+    id: str = Field(description="Unique identifier for the node")
+    node_type: AssetNodeType = Field(description="Type of node: file or folder")
+    name: str = Field(description="Name of the file or folder")
+    parent_id: str | None = Field(default=None, description="Parent folder ID, None for root level")
+    order: int = Field(default=0, description="Sort order within parent folder, lower values first")
+    extension: str = Field(default="", description="File extension without dot, empty for folders")
+    size: int = Field(default=0, description="File size in bytes, 0 for folders")
+    checksum: str = Field(default="", description="SHA-256 checksum of file content, empty for folders")
+
+    @classmethod
+    def create_folder(cls, node_id: str, name: str, parent_id: str | None = None) -> AppAssetNode:
+        return cls(id=node_id, node_type=AssetNodeType.FOLDER, name=name, parent_id=parent_id)
+
+    @classmethod
+    def create_file(
+        cls, node_id: str, name: str, parent_id: str | None = None, size: int = 0, checksum: str = ""
+    ) -> AppAssetNode:
+        return cls(
+            id=node_id,
+            node_type=AssetNodeType.FILE,
+            name=name,
+            parent_id=parent_id,
+            extension=name.rsplit(".", 1)[-1] if "." in name else "",
+            size=size,
+            checksum=checksum,
+        )
+
+
+class AppAssetNodeView(BaseModel):
+    id: str = Field(description="Unique identifier for the node")
+    node_type: str = Field(description="Type of node: 'file' or 'folder'")
+    name: str = Field(description="Name of the file or folder")
+    path: str = Field(description="Full path from root, e.g. '/folder/file.txt'")
+    extension: str = Field(default="", description="File extension without dot")
+    size: int = Field(default=0, description="File size in bytes")
+    checksum: str = Field(default="", description="SHA-256 checksum of file content")
+    children: list[AppAssetNodeView] = Field(default_factory=list, description="Child nodes for folders")
+
+
+class TreeNodeNotFoundError(Exception):
+    """Tree internal: node not found"""
+
+    pass
+
+
+class TreeParentNotFoundError(Exception):
+    """Tree internal: parent folder not found"""
+
+    pass
+
+
+class TreePathConflictError(Exception):
+    """Tree internal: path already exists"""
+
+    pass
+
+
+class AppAssetFileTree(BaseModel):
+    """
+    File tree structure for app assets using adjacency list pattern.
+
+    Design:
+    - Storage: Flat list with parent_id references (adjacency list)
+    - Path: Computed dynamically via get_path(), not stored
+    - Order: Integer field for user-defined sorting within each folder
+    - API response: transform() builds nested tree with computed paths
+
+    Why adjacency list over nested tree or materialized path:
+    - Simpler CRUD: move/rename only updates one node's parent_id
+    - No path cascade: renaming parent doesn't require updating all descendants
+    - JSON-friendly: flat list serializes cleanly to database JSON column
+    - Trade-off: path lookup is O(depth), acceptable for typical file trees
+    """
+
+    nodes: list[AppAssetNode] = Field(default_factory=list, description="Flat list of all nodes in the tree")
+
+    def get(self, node_id: str) -> AppAssetNode | None:
+        return next((n for n in self.nodes if n.id == node_id), None)
+
+    def get_children(self, parent_id: str | None) -> list[AppAssetNode]:
+        return [n for n in self.nodes if n.parent_id == parent_id]
+
+    def has_child_named(self, parent_id: str | None, name: str) -> bool:
+        return any(n.name == name and n.parent_id == parent_id for n in self.nodes)
+
+    def get_path(self, node_id: str) -> str:
+        node = self.get(node_id)
+        if not node:
+            raise TreeNodeNotFoundError(node_id)
+        parts: list[str] = []
+        current: AppAssetNode | None = node
+        while current:
+            parts.append(current.name)
+            current = self.get(current.parent_id) if current.parent_id else None
+        return "/".join(reversed(parts))
+
+    def relative_path(self, a: AppAssetNode, b: AppAssetNode) -> str:
+        """
+        Calculate relative path from node a to node b for Markdown references.
+        Path is computed from a's parent directory (where the file resides).
+
+        Examples:
+            /foo/a.md -> /foo/b.md         => ./b.md
+            /foo/a.md -> /foo/sub/b.md     => ./sub/b.md
+            /foo/sub/a.md -> /foo/b.md     => ../b.md
+            /foo/sub/deep/a.md -> /foo/b.md => ../../b.md
+        """
+
+        def get_ancestor_ids(node_id: str | None) -> list[str]:
+            chain: list[str] = []
+            current_id = node_id
+            while current_id:
+                chain.append(current_id)
+                node = self.get(current_id)
+                current_id = node.parent_id if node else None
+            return chain
+
+        a_dir_ancestors = get_ancestor_ids(a.parent_id)
+        b_ancestors = [b.id] + get_ancestor_ids(b.parent_id)
+        a_dir_set = set(a_dir_ancestors)
+
+        lca_id: str | None = None
+        lca_index_in_b = -1
+        for idx, ancestor_id in enumerate(b_ancestors):
+            if ancestor_id in a_dir_set or (a.parent_id is None and b_ancestors[idx:] == []):
+                lca_id = ancestor_id
+                lca_index_in_b = idx
+                break
+
+        if a.parent_id is None:
+            steps_up = 0
+            lca_index_in_b = len(b_ancestors)
+        elif lca_id is None:
+            steps_up = len(a_dir_ancestors)
+            lca_index_in_b = len(b_ancestors)
+        else:
+            steps_up = 0
+            for ancestor_id in a_dir_ancestors:
+                if ancestor_id == lca_id:
+                    break
+                steps_up += 1
+
+        path_down: list[str] = []
+        for i in range(lca_index_in_b - 1, -1, -1):
+            node = self.get(b_ancestors[i])
+            if node:
+                path_down.append(node.name)
+
+        if steps_up == 0:
+            return "./" + "/".join(path_down)
+
+        parts: list[str] = [".."] * steps_up + path_down
+        return "/".join(parts)
+
+    def get_descendant_ids(self, node_id: str) -> list[str]:
+        result: list[str] = []
+        stack = [node_id]
+        while stack:
+            current_id = stack.pop()
+            for child in self.nodes:
+                if child.parent_id == current_id:
+                    result.append(child.id)
+                    stack.append(child.id)
+        return result
+
+    def add(self, node: AppAssetNode) -> AppAssetNode:
+        if self.get(node.id):
+            raise TreePathConflictError(node.id)
+        if self.has_child_named(node.parent_id, node.name):
+            raise TreePathConflictError(node.name)
+        if node.parent_id:
+            parent = self.get(node.parent_id)
+            if not parent or parent.node_type != AssetNodeType.FOLDER:
+                raise TreeParentNotFoundError(node.parent_id)
+        siblings = self.get_children(node.parent_id)
+        node.order = max((s.order for s in siblings), default=-1) + 1
+        self.nodes.append(node)
+        return node
+
+    def update(self, node_id: str, size: int, checksum: str) -> AppAssetNode:
+        node = self.get(node_id)
+        if not node or node.node_type != AssetNodeType.FILE:
+            raise TreeNodeNotFoundError(node_id)
+        node.size = size
+        node.checksum = checksum
+        return node
+
+    def rename(self, node_id: str, new_name: str) -> AppAssetNode:
+        node = self.get(node_id)
+        if not node:
+            raise TreeNodeNotFoundError(node_id)
+        if node.name != new_name and self.has_child_named(node.parent_id, new_name):
+            raise TreePathConflictError(new_name)
+        node.name = new_name
+        if node.node_type == AssetNodeType.FILE:
+            node.extension = new_name.rsplit(".", 1)[-1] if "." in new_name else ""
+        return node
+
+    def move(self, node_id: str, new_parent_id: str | None) -> AppAssetNode:
+        node = self.get(node_id)
+        if not node:
+            raise TreeNodeNotFoundError(node_id)
+        if new_parent_id:
+            parent = self.get(new_parent_id)
+            if not parent or parent.node_type != AssetNodeType.FOLDER:
+                raise TreeParentNotFoundError(new_parent_id)
+        if self.has_child_named(new_parent_id, node.name):
+            raise TreePathConflictError(node.name)
+        node.parent_id = new_parent_id
+        siblings = self.get_children(new_parent_id)
+        node.order = max((s.order for s in siblings if s.id != node_id), default=-1) + 1
+        return node
+
+    def reorder(self, node_id: str, after_node_id: str | None) -> AppAssetNode:
+        node = self.get(node_id)
+        if not node:
+            raise TreeNodeNotFoundError(node_id)
+
+        siblings = sorted(self.get_children(node.parent_id), key=lambda x: x.order)
+        siblings = [s for s in siblings if s.id != node_id]
+
+        if after_node_id is None:
+            insert_idx = 0
+        else:
+            after_node = self.get(after_node_id)
+            if not after_node or after_node.parent_id != node.parent_id:
+                raise TreeNodeNotFoundError(after_node_id)
+            insert_idx = next((i for i, s in enumerate(siblings) if s.id == after_node_id), -1) + 1
+
+        siblings.insert(insert_idx, node)
+        for idx, sibling in enumerate(siblings):
+            sibling.order = idx
+
+        return node
+
+    def remove(self, node_id: str) -> list[str]:
+        node = self.get(node_id)
+        if not node:
+            raise TreeNodeNotFoundError(node_id)
+        ids_to_remove = [node_id] + self.get_descendant_ids(node_id)
+        self.nodes = [n for n in self.nodes if n.id not in ids_to_remove]
+        return ids_to_remove
+
+    def walk_files(self) -> Generator[AppAssetNode, None, None]:
+        return (n for n in self.nodes if n.node_type == AssetNodeType.FILE)
+
+    def transform(self) -> list[AppAssetNodeView]:
+        by_parent: dict[str | None, list[AppAssetNode]] = defaultdict(list)
+        for n in self.nodes:
+            by_parent[n.parent_id].append(n)
+
+        for children in by_parent.values():
+            children.sort(key=lambda x: x.order)
+
+        paths: dict[str, str] = {}
+        tree_views: dict[str, AppAssetNodeView] = {}
+
+        def build_view(node: AppAssetNode, parent_path: str) -> None:
+            path = f"{parent_path}/{node.name}"
+            paths[node.id] = path
+            child_views: list[AppAssetNodeView] = []
+            for child in by_parent.get(node.id, []):
+                build_view(child, path)
+                child_views.append(tree_views[child.id])
+            tree_views[node.id] = AppAssetNodeView(
+                id=node.id,
+                node_type=node.node_type.value,
+                name=node.name,
+                path=path,
+                extension=node.extension,
+                size=node.size,
+                checksum=node.checksum,
+                children=child_views,
+            )
+
+        for root_node in by_parent.get(None, []):
+            build_view(root_node, "")
+
+        return [tree_views[n.id] for n in by_parent.get(None, [])]

api/core/app/entities/app_invoke_entities.py

@@ -36,6 +36,9 @@ class InvokeFrom(StrEnum):
     # this is used for plugin trigger and webhook trigger.
     TRIGGER = "trigger"
 
+    # AGENT indicates that this invocation is from an agent.
+    AGENT = "agent"
+
     # EXPLORE indicates that this invocation is from
     # the workflow (or chatflow) explore page.
     EXPLORE = "explore"

api/core/app/entities/llm_generation_entities.py

@@ -0,0 +1,72 @@
+"""
+LLM Generation Detail entities.
+
+Defines the structure for storing and transmitting LLM generation details
+including reasoning content, tool calls, and their sequence.
+"""
+
+from typing import Literal
+
+from pydantic import BaseModel, Field
+
+
+class ContentSegment(BaseModel):
+    """Represents a content segment in the generation sequence."""
+
+    type: Literal["content"] = "content"
+    start: int = Field(..., description="Start position in the text")
+    end: int = Field(..., description="End position in the text")
+
+
+class ReasoningSegment(BaseModel):
+    """Represents a reasoning segment in the generation sequence."""
+
+    type: Literal["reasoning"] = "reasoning"
+    index: int = Field(..., description="Index into reasoning_content array")
+
+
+class ToolCallSegment(BaseModel):
+    """Represents a tool call segment in the generation sequence."""
+
+    type: Literal["tool_call"] = "tool_call"
+    index: int = Field(..., description="Index into tool_calls array")
+
+
+SequenceSegment = ContentSegment | ReasoningSegment | ToolCallSegment
+
+
+class ToolCallDetail(BaseModel):
+    """Represents a tool call with its arguments and result."""
+
+    id: str = Field(default="", description="Unique identifier for the tool call")
+    name: str = Field(..., description="Name of the tool")
+    arguments: str = Field(default="", description="JSON string of tool arguments")
+    result: str = Field(default="", description="Result from the tool execution")
+    elapsed_time: float | None = Field(default=None, description="Elapsed time in seconds")
+    icon: str | dict | None = Field(default=None, description="Icon of the tool")
+    icon_dark: str | dict | None = Field(default=None, description="Dark theme icon of the tool")
+
+
+class LLMGenerationDetailData(BaseModel):
+    """
+    Domain model for LLM generation detail.
+
+    Contains the structured data for reasoning content, tool calls,
+    and their display sequence.
+    """
+
+    reasoning_content: list[str] = Field(default_factory=list, description="List of reasoning segments")
+    tool_calls: list[ToolCallDetail] = Field(default_factory=list, description="List of tool call details")
+    sequence: list[SequenceSegment] = Field(default_factory=list, description="Display order of segments")
+
+    def is_empty(self) -> bool:
+        """Check if there's any meaningful generation detail."""
+        return not self.reasoning_content and not self.tool_calls
+
+    def to_response_dict(self) -> dict:
+        """Convert to dictionary for API response."""
+        return {
+            "reasoning_content": self.reasoning_content,
+            "tool_calls": [tc.model_dump() for tc in self.tool_calls],
+            "sequence": [seg.model_dump() for seg in self.sequence],
+        }

api/core/app/entities/queue_entities.py

@@ -7,7 +7,7 @@ from pydantic import BaseModel, ConfigDict, Field
 
 from core.model_runtime.entities.llm_entities import LLMResult, LLMResultChunk
 from core.rag.entities.citation_metadata import RetrievalSourceMetadata
-from core.workflow.entities import AgentNodeStrategyInit
+from core.workflow.entities import AgentNodeStrategyInit, ToolCall, ToolResult
 from core.workflow.enums import WorkflowNodeExecutionMetadataKey
 from core.workflow.nodes import NodeType
 
@@ -177,6 +177,17 @@ class QueueLoopCompletedEvent(AppQueueEvent):
     error: str | None = None
 
 
+class ChunkType(StrEnum):
+    """Stream chunk type for LLM-related events."""
+
+    TEXT = "text"  # Normal text streaming
+    TOOL_CALL = "tool_call"  # Tool call arguments streaming
+    TOOL_RESULT = "tool_result"  # Tool execution result
+    THOUGHT = "thought"  # Agent thinking process (ReAct)
+    THOUGHT_START = "thought_start"  # Agent thought start
+    THOUGHT_END = "thought_end"  # Agent thought end
+
+
 class QueueTextChunkEvent(AppQueueEvent):
     """
     QueueTextChunkEvent entity
@@ -190,6 +201,18 @@ class QueueTextChunkEvent(AppQueueEvent):
     """iteration id if node is in iteration"""
     in_loop_id: str | None = None
     """loop id if node is in loop"""
+    in_parent_node_id: str | None = None
+    """parent node id if this is an extractor node event"""
+
+    # Extended fields for Agent/Tool streaming
+    chunk_type: ChunkType = ChunkType.TEXT
+    """type of the chunk"""
+
+    # Tool streaming payloads
+    tool_call: ToolCall | None = None
+    """structured tool call info"""
+    tool_result: ToolResult | None = None
+    """structured tool result info"""
 
 
 class QueueAgentMessageEvent(AppQueueEvent):
@@ -229,6 +252,8 @@ class QueueRetrieverResourcesEvent(AppQueueEvent):
     """iteration id if node is in iteration"""
     in_loop_id: str | None = None
     """loop id if node is in loop"""
+    in_parent_node_id: str | None = None
+    """parent node id if this is an extractor node event"""
 
 
 class QueueAnnotationReplyEvent(AppQueueEvent):
@@ -306,6 +331,8 @@ class QueueNodeStartedEvent(AppQueueEvent):
     node_run_index: int = 1  # FIXME(-LAN-): may not used
     in_iteration_id: str | None = None
     in_loop_id: str | None = None
+    in_parent_node_id: str | None = None
+    """parent node id if this is an extractor node event"""
     start_at: datetime
     agent_strategy: AgentNodeStrategyInit | None = None
 
@@ -328,6 +355,8 @@ class QueueNodeSucceededEvent(AppQueueEvent):
     """iteration id if node is in iteration"""
     in_loop_id: str | None = None
     """loop id if node is in loop"""
+    in_parent_node_id: str | None = None
+    """parent node id if this is an extractor node event"""
     start_at: datetime
 
     inputs: Mapping[str, object] = Field(default_factory=dict)
@@ -383,6 +412,8 @@ class QueueNodeExceptionEvent(AppQueueEvent):
     """iteration id if node is in iteration"""
     in_loop_id: str | None = None
     """loop id if node is in loop"""
+    in_parent_node_id: str | None = None
+    """parent node id if this is an extractor node event"""
     start_at: datetime
 
     inputs: Mapping[str, object] = Field(default_factory=dict)
@@ -407,6 +438,8 @@ class QueueNodeFailedEvent(AppQueueEvent):
     """iteration id if node is in iteration"""
     in_loop_id: str | None = None
     """loop id if node is in loop"""
+    in_parent_node_id: str | None = None
+    """parent node id if this is an extractor node event"""
     start_at: datetime
 
     inputs: Mapping[str, object] = Field(default_factory=dict)

api/core/app/entities/task_entities.py

@@ -113,6 +113,38 @@ class MessageStreamResponse(StreamResponse):
     answer: str
     from_variable_selector: list[str] | None = None
 
+    # Extended fields for Agent/Tool streaming (imported at runtime to avoid circular import)
+    chunk_type: str | None = None
+    """type of the chunk: text, tool_call, tool_result, thought"""
+
+    # Tool call fields (when chunk_type == "tool_call")
+    tool_call_id: str | None = None
+    """unique identifier for this tool call"""
+    tool_name: str | None = None
+    """name of the tool being called"""
+    tool_arguments: str | None = None
+    """accumulated tool arguments JSON"""
+
+    # Tool result fields (when chunk_type == "tool_result")
+    tool_files: list[str] | None = None
+    """file IDs produced by tool"""
+    tool_error: str | None = None
+    """error message if tool failed"""
+    tool_elapsed_time: float | None = None
+    """elapsed time spent executing the tool"""
+    tool_icon: str | dict | None = None
+    """icon of the tool"""
+    tool_icon_dark: str | dict | None = None
+    """dark theme icon of the tool"""
+
+    def model_dump(self, *args, **kwargs) -> dict[str, object]:
+        kwargs.setdefault("exclude_none", True)
+        return super().model_dump(*args, **kwargs)
+
+    def model_dump_json(self, *args, **kwargs) -> str:
+        kwargs.setdefault("exclude_none", True)
+        return super().model_dump_json(*args, **kwargs)
+
 
 class MessageAudioStreamResponse(StreamResponse):
     """
@@ -262,6 +294,7 @@ class NodeStartStreamResponse(StreamResponse):
         extras: dict[str, object] = Field(default_factory=dict)
         iteration_id: str | None = None
         loop_id: str | None = None
+        parent_node_id: str | None = None
         agent_strategy: AgentNodeStrategyInit | None = None
 
     event: StreamEvent = StreamEvent.NODE_STARTED
@@ -285,6 +318,7 @@ class NodeStartStreamResponse(StreamResponse):
                 "extras": {},
                 "iteration_id": self.data.iteration_id,
                 "loop_id": self.data.loop_id,
+                "parent_node_id": self.data.parent_node_id,
             },
         }
 
@@ -320,6 +354,7 @@ class NodeFinishStreamResponse(StreamResponse):
         files: Sequence[Mapping[str, Any]] | None = []
         iteration_id: str | None = None
         loop_id: str | None = None
+        parent_node_id: str | None = None
 
     event: StreamEvent = StreamEvent.NODE_FINISHED
     workflow_run_id: str
@@ -349,6 +384,7 @@ class NodeFinishStreamResponse(StreamResponse):
                 "files": [],
                 "iteration_id": self.data.iteration_id,
                 "loop_id": self.data.loop_id,
+                "parent_node_id": self.data.parent_node_id,
             },
         }
 
@@ -384,6 +420,7 @@ class NodeRetryStreamResponse(StreamResponse):
         files: Sequence[Mapping[str, Any]] | None = []
         iteration_id: str | None = None
         loop_id: str | None = None
+        parent_node_id: str | None = None
         retry_index: int = 0
 
     event: StreamEvent = StreamEvent.NODE_RETRY
@@ -414,6 +451,7 @@ class NodeRetryStreamResponse(StreamResponse):
                 "files": [],
                 "iteration_id": self.data.iteration_id,
                 "loop_id": self.data.loop_id,
+                "parent_node_id": self.data.parent_node_id,
                 "retry_index": self.data.retry_index,
             },
         }
@@ -582,6 +620,17 @@ class LoopNodeCompletedStreamResponse(StreamResponse):
     data: Data
 
 
+class ChunkType(StrEnum):
+    """Stream chunk type for LLM-related events."""
+
+    TEXT = "text"  # Normal text streaming
+    TOOL_CALL = "tool_call"  # Tool call arguments streaming
+    TOOL_RESULT = "tool_result"  # Tool execution result
+    THOUGHT = "thought"  # Agent thinking process (ReAct)
+    THOUGHT_START = "thought_start"  # Agent thought start
+    THOUGHT_END = "thought_end"  # Agent thought end
+
+
 class TextChunkStreamResponse(StreamResponse):
     """
     TextChunkStreamResponse entity
@@ -595,6 +644,36 @@ class TextChunkStreamResponse(StreamResponse):
         text: str
         from_variable_selector: list[str] | None = None
 
+        # Extended fields for Agent/Tool streaming
+        chunk_type: ChunkType = ChunkType.TEXT
+        """type of the chunk"""
+
+        # Tool call fields (when chunk_type == TOOL_CALL)
+        tool_call_id: str | None = None
+        """unique identifier for this tool call"""
+        tool_name: str | None = None
+        """name of the tool being called"""
+        tool_arguments: str | None = None
+        """accumulated tool arguments JSON"""
+
+        # Tool result fields (when chunk_type == TOOL_RESULT)
+        tool_files: list[str] | None = None
+        """file IDs produced by tool"""
+        tool_error: str | None = None
+        """error message if tool failed"""
+
+        # Tool elapsed time fields (when chunk_type == TOOL_RESULT)
+        tool_elapsed_time: float | None = None
+        """elapsed time spent executing the tool"""
+
+        def model_dump(self, *args, **kwargs) -> dict[str, object]:
+            kwargs.setdefault("exclude_none", True)
+            return super().model_dump(*args, **kwargs)
+
+        def model_dump_json(self, *args, **kwargs) -> str:
+            kwargs.setdefault("exclude_none", True)
+            return super().model_dump_json(*args, **kwargs)
+
     event: StreamEvent = StreamEvent.TEXT_CHUNK
     data: Data
 
@@ -743,7 +822,7 @@ class AgentLogStreamResponse(StreamResponse):
         """
 
         node_execution_id: str
-        id: str
+        message_id: str
         label: str
         parent_id: str | None = None
         error: str | None = None

api/core/app/layers/sandbox_layer.py

@@ -0,0 +1,22 @@
+import logging
+
+from core.sandbox import Sandbox
+from core.workflow.graph_engine.layers.base import GraphEngineLayer
+from core.workflow.graph_events.base import GraphEngineEvent
+
+logger = logging.getLogger(__name__)
+
+
+class SandboxLayer(GraphEngineLayer):
+    def __init__(self, sandbox: Sandbox) -> None:
+        super().__init__()
+        self._sandbox = sandbox
+
+    def on_graph_start(self) -> None:
+        pass
+
+    def on_event(self, event: GraphEngineEvent) -> None:
+        pass
+
+    def on_graph_end(self, error: Exception | None) -> None:
+        self._sandbox.release()

api/core/app/task_pipeline/easy_ui_based_generate_task_pipeline.py

@@ -1,4 +1,5 @@
 import logging
+import re
 import time
 from collections.abc import Generator
 from threading import Thread
@@ -58,7 +59,7 @@ from core.prompt.utils.prompt_template_parser import PromptTemplateParser
 from events.message_event import message_was_created
 from extensions.ext_database import db
 from libs.datetime_utils import naive_utc_now
-from models.model import AppMode, Conversation, Message, MessageAgentThought
+from models.model import AppMode, Conversation, LLMGenerationDetail, Message, MessageAgentThought
 
 logger = logging.getLogger(__name__)
 
@@ -68,6 +69,8 @@ class EasyUIBasedGenerateTaskPipeline(BasedGenerateTaskPipeline):
     EasyUIBasedGenerateTaskPipeline is a class that generate stream output and state management for Application.
     """
 
+    _THINK_PATTERN = re.compile(r"<think[^>]*>(.*?)</think>", re.IGNORECASE | re.DOTALL)
+
     _task_state: EasyUITaskState
     _application_generate_entity: Union[ChatAppGenerateEntity, CompletionAppGenerateEntity, AgentChatAppGenerateEntity]
 
@@ -409,11 +412,136 @@ class EasyUIBasedGenerateTaskPipeline(BasedGenerateTaskPipeline):
                 )
             )
 
+        # Save LLM generation detail if there's reasoning_content
+        self._save_generation_detail(session=session, message=message, llm_result=llm_result)
+
         message_was_created.send(
             message,
             application_generate_entity=self._application_generate_entity,
         )
 
+    def _save_generation_detail(self, *, session: Session, message: Message, llm_result: LLMResult) -> None:
+        """
+        Save LLM generation detail for Completion/Chat/Agent-Chat applications.
+        For Agent-Chat, also merges MessageAgentThought records.
+        """
+        import json
+
+        reasoning_list: list[str] = []
+        tool_calls_list: list[dict] = []
+        sequence: list[dict] = []
+        answer = message.answer or ""
+
+        # Check if this is Agent-Chat mode by looking for agent thoughts
+        agent_thoughts = (
+            session.query(MessageAgentThought)
+            .filter_by(message_id=message.id)
+            .order_by(MessageAgentThought.position.asc())
+            .all()
+        )
+
+        if agent_thoughts:
+            # Agent-Chat mode: merge MessageAgentThought records
+            content_pos = 0
+            cleaned_answer_parts: list[str] = []
+            for thought in agent_thoughts:
+                # Add thought/reasoning
+                if thought.thought:
+                    reasoning_text = thought.thought
+                    if "<think" in reasoning_text.lower():
+                        clean_text, extracted_reasoning = self._split_reasoning_from_answer(reasoning_text)
+                        if extracted_reasoning:
+                            reasoning_text = extracted_reasoning
+                            thought.thought = clean_text or extracted_reasoning
+                    reasoning_list.append(reasoning_text)
+                    sequence.append({"type": "reasoning", "index": len(reasoning_list) - 1})
+
+                # Add tool calls
+                if thought.tool:
+                    tool_calls_list.append(
+                        {
+                            "name": thought.tool,
+                            "arguments": thought.tool_input or "",
+                            "result": thought.observation or "",
+                        }
+                    )
+                    sequence.append({"type": "tool_call", "index": len(tool_calls_list) - 1})
+
+                # Add answer content if present
+                if thought.answer:
+                    content_text = thought.answer
+                    if "<think" in content_text.lower():
+                        clean_answer, extracted_reasoning = self._split_reasoning_from_answer(content_text)
+                        if extracted_reasoning:
+                            reasoning_list.append(extracted_reasoning)
+                            sequence.append({"type": "reasoning", "index": len(reasoning_list) - 1})
+                        content_text = clean_answer
+                        thought.answer = clean_answer or content_text
+
+                    if content_text:
+                        start = content_pos
+                        end = content_pos + len(content_text)
+                        sequence.append({"type": "content", "start": start, "end": end})
+                        content_pos = end
+                        cleaned_answer_parts.append(content_text)
+
+            if cleaned_answer_parts:
+                merged_answer = "".join(cleaned_answer_parts)
+                message.answer = merged_answer
+                llm_result.message.content = merged_answer
+        else:
+            # Completion/Chat mode: use reasoning_content from llm_result
+            reasoning_content = llm_result.reasoning_content
+            if not reasoning_content and answer:
+                # Extract reasoning from <think> blocks and clean the final answer
+                clean_answer, reasoning_content = self._split_reasoning_from_answer(answer)
+                if reasoning_content:
+                    answer = clean_answer
+                    llm_result.message.content = clean_answer
+                    llm_result.reasoning_content = reasoning_content
+                    message.answer = clean_answer
+            if reasoning_content:
+                reasoning_list = [reasoning_content]
+                # Content comes first, then reasoning
+                if answer:
+                    sequence.append({"type": "content", "start": 0, "end": len(answer)})
+                sequence.append({"type": "reasoning", "index": 0})
+
+        # Only save if there's meaningful generation detail
+        if not reasoning_list and not tool_calls_list:
+            return
+
+        # Check if generation detail already exists
+        existing = session.query(LLMGenerationDetail).filter_by(message_id=message.id).first()
+
+        if existing:
+            existing.reasoning_content = json.dumps(reasoning_list) if reasoning_list else None
+            existing.tool_calls = json.dumps(tool_calls_list) if tool_calls_list else None
+            existing.sequence = json.dumps(sequence) if sequence else None
+        else:
+            generation_detail = LLMGenerationDetail(
+                tenant_id=self._application_generate_entity.app_config.tenant_id,
+                app_id=self._application_generate_entity.app_config.app_id,
+                message_id=message.id,
+                reasoning_content=json.dumps(reasoning_list) if reasoning_list else None,
+                tool_calls=json.dumps(tool_calls_list) if tool_calls_list else None,
+                sequence=json.dumps(sequence) if sequence else None,
+            )
+            session.add(generation_detail)
+
+    @classmethod
+    def _split_reasoning_from_answer(cls, text: str) -> tuple[str, str]:
+        """
+        Extract reasoning segments from <think> blocks and return (clean_text, reasoning).
+        """
+        matches = cls._THINK_PATTERN.findall(text)
+        reasoning_content = "\n".join(match.strip() for match in matches) if matches else ""
+
+        clean_text = cls._THINK_PATTERN.sub("", text)
+        clean_text = re.sub(r"\n\s*\n", "\n\n", clean_text).strip()
+
+        return clean_text, reasoning_content or ""
+
     def _handle_stop(self, event: QueueStopEvent):
         """
         Handle stop.

api/core/app/task_pipeline/message_cycle_manager.py

@@ -232,15 +232,31 @@ class MessageCycleManager:
         answer: str,
         message_id: str,
         from_variable_selector: list[str] | None = None,
+        chunk_type: str | None = None,
+        tool_call_id: str | None = None,
+        tool_name: str | None = None,
+        tool_arguments: str | None = None,
+        tool_files: list[str] | None = None,
+        tool_error: str | None = None,
+        tool_elapsed_time: float | None = None,
+        tool_icon: str | dict | None = None,
+        tool_icon_dark: str | dict | None = None,
         event_type: StreamEvent | None = None,
     ) -> MessageStreamResponse:
         """
         Message to stream response.
         :param answer: answer
         :param message_id: message id
+        :param from_variable_selector: from variable selector
+        :param chunk_type: type of the chunk (text, function_call, tool_result, thought)
+        :param tool_call_id: unique identifier for this tool call
+        :param tool_name: name of the tool being called
+        :param tool_arguments: accumulated tool arguments JSON
+        :param tool_files: file IDs produced by tool
+        :param tool_error: error message if tool failed
         :return:
         """
-        return MessageStreamResponse(
+        response = MessageStreamResponse(
             task_id=self._application_generate_entity.task_id,
             id=message_id,
             answer=answer,
@@ -248,6 +264,35 @@ class MessageCycleManager:
             event=event_type or StreamEvent.MESSAGE,
         )
 
+        if chunk_type:
+            response = response.model_copy(update={"chunk_type": chunk_type})
+
+        if chunk_type == "tool_call":
+            response = response.model_copy(
+                update={
+                    "tool_call_id": tool_call_id,
+                    "tool_name": tool_name,
+                    "tool_arguments": tool_arguments,
+                    "tool_icon": tool_icon,
+                    "tool_icon_dark": tool_icon_dark,
+                }
+            )
+        elif chunk_type == "tool_result":
+            response = response.model_copy(
+                update={
+                    "tool_call_id": tool_call_id,
+                    "tool_name": tool_name,
+                    "tool_arguments": tool_arguments,
+                    "tool_files": tool_files,
+                    "tool_error": tool_error,
+                    "tool_elapsed_time": tool_elapsed_time,
+                    "tool_icon": tool_icon,
+                    "tool_icon_dark": tool_icon_dark,
+                }
+            )
+
+        return response
+
     def message_replace_to_stream_response(self, answer: str, reason: str = "") -> MessageReplaceStreamResponse:
         """
         Message replace to stream response.

api/core/app_assets/__init__.py

@@ -0,0 +1,23 @@
+from .constants import AppAssetsAttrs
+from .entities import (
+    AssetItem,
+    FileAsset,
+    SkillAsset,
+)
+from .packager import AssetPackager, ZipPackager
+from .parser import AssetItemParser, AssetParser, FileAssetParser, SkillAssetParser
+from .paths import AssetPaths
+
+__all__ = [
+    "AppAssetsAttrs",
+    "AssetItem",
+    "AssetItemParser",
+    "AssetPackager",
+    "AssetParser",
+    "AssetPaths",
+    "FileAsset",
+    "FileAssetParser",
+    "SkillAsset",
+    "SkillAssetParser",
+    "ZipPackager",
+]

api/core/app_assets/builder/__init__.py

@@ -0,0 +1,12 @@
+from .base import AssetBuilder, BuildContext
+from .file_builder import FileBuilder
+from .pipeline import AssetBuildPipeline
+from .skill_builder import SkillBuilder
+
+__all__ = [
+    "AssetBuildPipeline",
+    "AssetBuilder",
+    "BuildContext",
+    "FileBuilder",
+    "SkillBuilder",
+]

api/core/app_assets/builder/base.py

@@ -0,0 +1,20 @@
+from dataclasses import dataclass
+from typing import Protocol
+
+from core.app.entities.app_asset_entities import AppAssetFileTree, AppAssetNode
+from core.app_assets.entities import AssetItem
+
+
+@dataclass
+class BuildContext:
+    tenant_id: str
+    app_id: str
+    build_id: str
+
+
+class AssetBuilder(Protocol):
+    def accept(self, node: AppAssetNode) -> bool: ...
+
+    def collect(self, node: AppAssetNode, path: str, ctx: BuildContext) -> None: ...
+
+    def build(self, tree: AppAssetFileTree, ctx: BuildContext) -> list[AssetItem]: ...

api/core/app_assets/builder/file_builder.py

@@ -0,0 +1,30 @@
+from core.app.entities.app_asset_entities import AppAssetFileTree, AppAssetNode
+from core.app_assets.entities import AssetItem, FileAsset
+from core.app_assets.paths import AssetPaths
+
+from .base import BuildContext
+
+
+class FileBuilder:
+    _nodes: list[tuple[AppAssetNode, str]]
+
+    def __init__(self) -> None:
+        self._nodes = []
+
+    def accept(self, node: AppAssetNode) -> bool:
+        return True
+
+    def collect(self, node: AppAssetNode, path: str, ctx: BuildContext) -> None:
+        self._nodes.append((node, path))
+
+    def build(self, tree: AppAssetFileTree, ctx: BuildContext) -> list[AssetItem]:
+        return [
+            FileAsset(
+                asset_id=node.id,
+                path=path,
+                file_name=node.name,
+                extension=node.extension or "",
+                storage_key=AssetPaths.draft_file(ctx.tenant_id, ctx.app_id, node.id),
+            )
+            for node, path in self._nodes
+        ]

api/core/app_assets/builder/pipeline.py

@@ -0,0 +1,29 @@
+from core.app.entities.app_asset_entities import AppAssetFileTree
+from core.app_assets.builder.file_builder import FileBuilder
+from core.app_assets.builder.skill_builder import SkillBuilder
+from core.app_assets.entities import AssetItem
+
+from .base import AssetBuilder, BuildContext
+
+
+class AssetBuildPipeline:
+    _builders: list[AssetBuilder]
+
+    def __init__(self, builders: list[AssetBuilder] | None = None) -> None:
+        self._builders = builders or [SkillBuilder(), FileBuilder()]
+
+    def build_all(self, tree: AppAssetFileTree, ctx: BuildContext) -> list[AssetItem]:
+        # 1. Distribute: each node goes to first accepting builder
+        for node in tree.walk_files():
+            path = tree.get_path(node.id)
+            for builder in self._builders:
+                if builder.accept(node):
+                    builder.collect(node, path, ctx)
+                    break
+
+        # 2. Each builder builds its collected nodes
+        results: list[AssetItem] = []
+        for builder in self._builders:
+            results.extend(builder.build(tree, ctx))
+
+        return results

api/core/app_assets/builder/skill_builder.py

@@ -0,0 +1,113 @@
+import json
+from concurrent.futures import ThreadPoolExecutor
+from dataclasses import dataclass
+
+from core.app.entities.app_asset_entities import AppAssetFileTree, AppAssetNode
+from core.app_assets.entities import AssetItem, FileAsset
+from core.app_assets.paths import AssetPaths
+from core.skill.entities.skill_document import SkillDocument
+from core.skill.skill_compiler import SkillCompiler
+from core.skill.skill_manager import SkillManager
+from extensions.ext_storage import storage
+
+from .base import BuildContext
+
+
+@dataclass
+class _LoadedSkill:
+    node: AppAssetNode
+    path: str
+    content: str
+    metadata: dict
+
+
+@dataclass
+class _CompiledSkill:
+    node: AppAssetNode
+    path: str
+    resolved_key: str
+    content_bytes: bytes
+
+
+class SkillBuilder:
+    _nodes: list[tuple[AppAssetNode, str]]
+    _max_workers: int
+
+    def __init__(self, max_workers: int = 8) -> None:
+        self._nodes = []
+        self._max_workers = max_workers
+
+    def accept(self, node: AppAssetNode) -> bool:
+        return node.extension == "md"
+
+    def collect(self, node: AppAssetNode, path: str, ctx: BuildContext) -> None:
+        self._nodes.append((node, path))
+
+    def build(self, tree: AppAssetFileTree, ctx: BuildContext) -> list[AssetItem]:
+        if not self._nodes:
+            return []
+
+        # 1. Load all skills (parallel IO)
+        loaded = self._load_all(ctx)
+
+        # 2. Compile all skills (CPU-bound, single thread)
+        documents = [SkillDocument(skill_id=s.node.id, content=s.content, metadata=s.metadata) for s in loaded]
+        artifact_set = SkillCompiler().compile_all(documents, tree, ctx.build_id)
+
+        SkillManager.save_bundle(ctx.tenant_id, ctx.app_id, ctx.build_id, artifact_set)
+
+        # 4. Prepare compiled skills for upload
+        to_upload: list[_CompiledSkill] = []
+        for skill in loaded:
+            artifact = artifact_set.get(skill.node.id)
+            if artifact is None:
+                continue
+            resolved_key = AssetPaths.build_resolved_file(ctx.tenant_id, ctx.app_id, ctx.build_id, skill.node.id)
+            to_upload.append(
+                _CompiledSkill(
+                    node=skill.node,
+                    path=skill.path,
+                    resolved_key=resolved_key,
+                    content_bytes=artifact.content.encode("utf-8"),
+                )
+            )
+
+        # 5. Upload all compiled skills (parallel IO)
+        self._upload_all(to_upload)
+
+        # 6. Return FileAssets
+        return [
+            FileAsset(
+                asset_id=s.node.id,
+                path=s.path,
+                file_name=s.node.name,
+                extension=s.node.extension or "",
+                storage_key=s.resolved_key,
+            )
+            for s in to_upload
+        ]
+
+    def _load_all(self, ctx: BuildContext) -> list[_LoadedSkill]:
+        def load_one(node: AppAssetNode, path: str) -> _LoadedSkill:
+            draft_key = AssetPaths.draft_file(ctx.tenant_id, ctx.app_id, node.id)
+            try:
+                data = json.loads(storage.load_once(draft_key))
+                content = data.get("content", "") if isinstance(data, dict) else ""
+                metadata = data.get("metadata", {}) if isinstance(data, dict) else {}
+            except Exception:
+                content = ""
+                metadata = {}
+            return _LoadedSkill(node=node, path=path, content=content, metadata=metadata)
+
+        with ThreadPoolExecutor(max_workers=self._max_workers) as executor:
+            futures = [executor.submit(load_one, node, path) for node, path in self._nodes]
+            return [f.result() for f in futures]
+
+    def _upload_all(self, skills: list[_CompiledSkill]) -> None:
+        def upload_one(skill: _CompiledSkill) -> None:
+            storage.save(skill.resolved_key, skill.content_bytes)
+
+        with ThreadPoolExecutor(max_workers=self._max_workers) as executor:
+            futures = [executor.submit(upload_one, skill) for skill in skills]
+            for f in futures:
+                f.result()

api/core/app_assets/constants.py

@@ -0,0 +1,7 @@
+from core.app.entities.app_asset_entities import AppAssetFileTree
+from libs.attr_map import AttrKey
+
+
+class AppAssetsAttrs:
+    # Skill artifact set
+    FILE_TREE = AttrKey("file_tree", AppAssetFileTree)

api/core/app_assets/entities/__init__.py

@@ -0,0 +1,8 @@
+from .assets import AssetItem, FileAsset
+from .skill import SkillAsset
+
+__all__ = [
+    "AssetItem",
+    "FileAsset",
+    "SkillAsset",
+]

api/core/app_assets/entities/assets.py

@@ -0,0 +1,22 @@
+from abc import ABC, abstractmethod
+from dataclasses import dataclass
+
+
+@dataclass
+class AssetItem(ABC):
+    asset_id: str
+    path: str
+    file_name: str
+    extension: str
+
+    @abstractmethod
+    def get_storage_key(self) -> str:
+        raise NotImplementedError
+
+
+@dataclass
+class FileAsset(AssetItem):
+    storage_key: str
+
+    def get_storage_key(self) -> str:
+        return self.storage_key

api/core/app_assets/entities/skill.py

@@ -0,0 +1,14 @@
+from collections.abc import Mapping
+from dataclasses import dataclass, field
+from typing import Any
+
+from .assets import AssetItem
+
+
+@dataclass
+class SkillAsset(AssetItem):
+    storage_key: str
+    metadata: Mapping[str, Any] = field(default_factory=dict)
+
+    def get_storage_key(self) -> str:
+        return self.storage_key

api/core/app_assets/packager/__init__.py

@@ -0,0 +1,7 @@
+from .base import AssetPackager
+from .zip_packager import ZipPackager
+
+__all__ = [
+    "AssetPackager",
+    "ZipPackager",
+]

api/core/app_assets/packager/base.py

@@ -0,0 +1,9 @@
+from abc import ABC, abstractmethod
+
+from core.app_assets.entities import AssetItem
+
+
+class AssetPackager(ABC):
+    @abstractmethod
+    def package(self, assets: list[AssetItem]) -> bytes:
+        raise NotImplementedError

api/core/app_assets/packager/zip_packager.py

@@ -0,0 +1,42 @@
+import io
+import zipfile
+from concurrent.futures import Future, ThreadPoolExecutor
+from threading import Lock
+from typing import TYPE_CHECKING
+
+from core.app_assets.entities import AssetItem
+
+from .base import AssetPackager
+
+if TYPE_CHECKING:
+    from extensions.ext_storage import Storage
+
+
+class ZipPackager(AssetPackager):
+    _storage: "Storage"
+
+    def __init__(self, storage: "Storage") -> None:
+        self._storage = storage
+
+    def package(self, assets: list[AssetItem]) -> bytes:
+        zip_buffer = io.BytesIO()
+
+        with zipfile.ZipFile(zip_buffer, "w", zipfile.ZIP_DEFLATED) as zf:
+            lock = Lock()
+            # FOR DELVELPMENT AND TESTING ONLY, TODO: optimize
+            with ThreadPoolExecutor(max_workers=8) as executor:
+                futures: list[Future[None]] = []
+                for asset in assets:
+
+                    def _write_asset(a: AssetItem) -> None:
+                        content = self._storage.load_once(a.get_storage_key())
+                        with lock:
+                            zf.writestr(a.path, content)
+
+                    futures.append(executor.submit(_write_asset, asset))
+
+                # Wait for all futures to complete
+                for future in futures:
+                    future.result()
+
+        return zip_buffer.getvalue()

api/core/app_assets/parser/__init__.py

@@ -0,0 +1,10 @@
+from .asset_parser import AssetParser
+from .base import AssetItemParser, FileAssetParser
+from .skill_parser import SkillAssetParser
+
+__all__ = [
+    "AssetItemParser",
+    "AssetParser",
+    "FileAssetParser",
+    "SkillAssetParser",
+]

api/core/app_assets/parser/asset_parser.py

@@ -0,0 +1,36 @@
+from core.app.entities.app_asset_entities import AppAssetFileTree
+from core.app_assets.entities import AssetItem
+from core.app_assets.paths import AssetPaths
+
+from .base import AssetItemParser, FileAssetParser
+
+
+class AssetParser:
+    def __init__(
+        self,
+        tree: AppAssetFileTree,
+        tenant_id: str,
+        app_id: str,
+    ) -> None:
+        self._tree = tree
+        self._tenant_id = tenant_id
+        self._app_id = app_id
+        self._parsers = {}
+        self._default_parser = FileAssetParser()
+
+    def register(self, extension: str, parser: AssetItemParser) -> None:
+        self._parsers[extension] = parser
+
+    def parse(self) -> list[AssetItem]:
+        assets: list[AssetItem] = []
+
+        for node in self._tree.walk_files():
+            path = self._tree.get_path(node.id).lstrip("/")
+            storage_key = AssetPaths.draft_file(self._tenant_id, self._app_id, node.id)
+            extension = node.extension or ""
+
+            parser = self._parsers.get(extension, self._default_parser)
+            asset = parser.parse(node.id, path, node.name, extension, storage_key)
+            assets.append(asset)
+
+        return assets

api/core/app_assets/parser/base.py

@@ -0,0 +1,34 @@
+from abc import ABC, abstractmethod
+
+from core.app_assets.entities import AssetItem, FileAsset
+
+
+class AssetItemParser(ABC):
+    @abstractmethod
+    def parse(
+        self,
+        asset_id: str,
+        path: str,
+        file_name: str,
+        extension: str,
+        storage_key: str,
+    ) -> AssetItem:
+        raise NotImplementedError
+
+
+class FileAssetParser(AssetItemParser):
+    def parse(
+        self,
+        asset_id: str,
+        path: str,
+        file_name: str,
+        extension: str,
+        storage_key: str,
+    ) -> FileAsset:
+        return FileAsset(
+            asset_id=asset_id,
+            path=path,
+            file_name=file_name,
+            extension=extension,
+            storage_key=storage_key,
+        )

api/core/app_assets/parser/skill_parser.py

@@ -0,0 +1,57 @@
+import json
+import logging
+from typing import Any
+
+from core.app_assets.entities import SkillAsset
+from core.app_assets.entities.assets import AssetItem, FileAsset
+from extensions.ext_storage import storage
+
+from .base import AssetItemParser
+
+logger = logging.getLogger(__name__)
+
+
+class SkillAssetParser(AssetItemParser):
+    """
+    Parser for skill assets.
+
+    Responsibilities:
+    - Read file from storage
+    - Parse JSON structure
+    - Return SkillAsset with raw metadata (no parsing/resolution)
+
+    Metadata parsing and content resolution are handled by SkillCompiler.
+    """
+
+    def parse(
+        self,
+        asset_id: str,
+        path: str,
+        file_name: str,
+        extension: str,
+        storage_key: str,
+    ) -> AssetItem:
+        try:
+            data = json.loads(storage.load_once(storage_key))
+            if not isinstance(data, dict):
+                raise ValueError(f"Skill document {asset_id} must be a JSON object")
+
+            metadata_raw: dict[str, Any] = data.get("metadata", {})
+
+            return SkillAsset(
+                asset_id=asset_id,
+                path=path,
+                file_name=file_name,
+                extension=extension,
+                storage_key=storage_key,
+                metadata=metadata_raw,
+            )
+        except Exception:
+            logger.exception("Failed to parse skill asset %s", asset_id)
+            return FileAsset(
+                asset_id=asset_id,
+                path=path,
+                file_name=file_name,
+                extension=extension,
+                storage_key=storage_key,
+            )

api/core/app_assets/paths.py

@@ -0,0 +1,18 @@
+class AssetPaths:
+    _BASE = "app_assets"
+
+    @staticmethod
+    def draft_file(tenant_id: str, app_id: str, node_id: str) -> str:
+        return f"{AssetPaths._BASE}/{tenant_id}/{app_id}/draft/{node_id}"
+
+    @staticmethod
+    def build_zip(tenant_id: str, app_id: str, assets_id: str) -> str:
+        return f"{AssetPaths._BASE}/{tenant_id}/{app_id}/artifacts/{assets_id}.zip"
+
+    @staticmethod
+    def build_resolved_file(tenant_id: str, app_id: str, assets_id: str, node_id: str) -> str:
+        return f"{AssetPaths._BASE}/{tenant_id}/{app_id}/artifacts/{assets_id}/resolved/{node_id}"
+
+    @staticmethod
+    def build_skill_artifact_set(tenant_id: str, app_id: str, assets_id: str) -> str:
+        return f"{AssetPaths._BASE}/{tenant_id}/{app_id}/artifacts/{assets_id}/skill_artifact_set.json"

api/core/callback_handler/index_tool_callback_handler.py

@@ -5,7 +5,6 @@ from sqlalchemy import select
 
 from core.app.apps.base_app_queue_manager import AppQueueManager, PublishFrom
 from core.app.entities.app_invoke_entities import InvokeFrom
-from core.app.entities.queue_entities import QueueRetrieverResourcesEvent
 from core.rag.entities.citation_metadata import RetrievalSourceMetadata
 from core.rag.index_processor.constant.index_type import IndexStructureType
 from core.rag.models.document import Document
@@ -90,6 +89,8 @@ class DatasetIndexToolCallbackHandler:
     # TODO(-LAN-): Improve type check
     def return_retriever_resource_info(self, resource: Sequence[RetrievalSourceMetadata]):
         """Handle return_retriever_resource_info."""
+        from core.app.entities.queue_entities import QueueRetrieverResourcesEvent
+
         self._queue_manager.publish(
             QueueRetrieverResourcesEvent(retriever_resources=resource), PublishFrom.APPLICATION_MANAGER
         )

api/core/file/file_manager.py

@@ -1,4 +1,5 @@
 import base64
+import logging
 from collections.abc import Mapping
 
 from configs import dify_config
@@ -10,7 +11,10 @@ from core.model_runtime.entities import (
     TextPromptMessageContent,
     VideoPromptMessageContent,
 )
-from core.model_runtime.entities.message_entities import PromptMessageContentUnionTypes
+from core.model_runtime.entities.message_entities import (
+    MultiModalPromptMessageContent,
+    PromptMessageContentUnionTypes,
+)
 from core.tools.signature import sign_tool_file
 from extensions.ext_storage import storage
 
@@ -18,6 +22,8 @@ from . import helpers
 from .enums import FileAttribute
 from .models import File, FileTransferMethod, FileType
 
+logger = logging.getLogger(__name__)
+
 
 def get_attr(*, file: File, attr: FileAttribute):
     match attr:
@@ -89,6 +95,8 @@ def to_prompt_message_content(
         "format": f.extension.removeprefix("."),
         "mime_type": f.mime_type,
         "filename": f.filename or "",
+        # Encoded file reference for context restoration: "transfer_method:related_id" or "remote:url"
+        "file_ref": _encode_file_ref(f),
     }
     if f.type == FileType.IMAGE:
         params["detail"] = image_detail_config or ImagePromptMessageContent.DETAIL.LOW
@@ -96,6 +104,17 @@ def to_prompt_message_content(
     return prompt_class_map[f.type].model_validate(params)
 
 
+def _encode_file_ref(f: File) -> str | None:
+    """Encode file reference as 'transfer_method:id_or_url' string."""
+    if f.transfer_method == FileTransferMethod.REMOTE_URL:
+        return f"remote:{f.remote_url}" if f.remote_url else None
+    elif f.transfer_method == FileTransferMethod.LOCAL_FILE:
+        return f"local:{f.related_id}" if f.related_id else None
+    elif f.transfer_method == FileTransferMethod.TOOL_FILE:
+        return f"tool:{f.related_id}" if f.related_id else None
+    return None
+
+
 def download(f: File, /):
     if f.transfer_method in (
         FileTransferMethod.TOOL_FILE,
@@ -164,3 +183,128 @@ def _to_url(f: File, /):
         return sign_tool_file(tool_file_id=f.related_id, extension=f.extension)
     else:
         raise ValueError(f"Unsupported transfer method: {f.transfer_method}")
+
+
+def restore_multimodal_content(
+    content: MultiModalPromptMessageContent,
+) -> MultiModalPromptMessageContent:
+    """
+    Restore base64_data or url for multimodal content from file_ref.
+
+    file_ref format: "transfer_method:id_or_url" (e.g., "local:abc123", "remote:https://...")
+
+    Args:
+        content: MultiModalPromptMessageContent with file_ref field
+
+    Returns:
+        MultiModalPromptMessageContent with restored base64_data or url
+    """
+    # Skip if no file reference or content already has data
+    if not content.file_ref:
+        return content
+    if content.base64_data or content.url:
+        return content
+
+    try:
+        file = _build_file_from_ref(
+            file_ref=content.file_ref,
+            file_format=content.format,
+            mime_type=content.mime_type,
+            filename=content.filename,
+        )
+        if not file:
+            return content
+
+        # Restore content based on config
+        if dify_config.MULTIMODAL_SEND_FORMAT == "base64":
+            restored_base64 = _get_encoded_string(file)
+            return content.model_copy(update={"base64_data": restored_base64})
+        else:
+            restored_url = _to_url(file)
+            return content.model_copy(update={"url": restored_url})
+
+    except Exception as e:
+        logger.warning("Failed to restore multimodal content: %s", e)
+        return content
+
+
+def _build_file_from_ref(
+    file_ref: str,
+    file_format: str | None,
+    mime_type: str | None,
+    filename: str | None,
+) -> File | None:
+    """
+    Build a File object from encoded file_ref string.
+
+    Args:
+        file_ref: Encoded reference "transfer_method:id_or_url"
+        file_format: The file format/extension (without dot)
+        mime_type: The mime type
+        filename: The filename
+
+    Returns:
+        File object with storage_key loaded, or None if not found
+    """
+    from sqlalchemy import select
+    from sqlalchemy.orm import Session
+
+    from extensions.ext_database import db
+    from models.model import UploadFile
+    from models.tools import ToolFile
+
+    # Parse file_ref: "method:value"
+    if ":" not in file_ref:
+        logger.warning("Invalid file_ref format: %s", file_ref)
+        return None
+
+    method, value = file_ref.split(":", 1)
+    extension = f".{file_format}" if file_format else None
+
+    if method == "remote":
+        return File(
+            tenant_id="",
+            type=FileType.IMAGE,
+            transfer_method=FileTransferMethod.REMOTE_URL,
+            remote_url=value,
+            extension=extension,
+            mime_type=mime_type,
+            filename=filename,
+            storage_key="",
+        )
+
+    # Query database for storage_key
+    with Session(db.engine) as session:
+        if method == "local":
+            stmt = select(UploadFile).where(UploadFile.id == value)
+            upload_file = session.scalar(stmt)
+            if upload_file:
+                return File(
+                    tenant_id=upload_file.tenant_id,
+                    type=FileType(upload_file.extension)
+                    if hasattr(FileType, upload_file.extension.upper())
+                    else FileType.IMAGE,
+                    transfer_method=FileTransferMethod.LOCAL_FILE,
+                    related_id=value,
+                    extension=extension or ("." + upload_file.extension if upload_file.extension else None),
+                    mime_type=mime_type or upload_file.mime_type,
+                    filename=filename or upload_file.name,
+                    storage_key=upload_file.key,
+                )
+        elif method == "tool":
+            stmt = select(ToolFile).where(ToolFile.id == value)
+            tool_file = session.scalar(stmt)
+            if tool_file:
+                return File(
+                    tenant_id=tool_file.tenant_id,
+                    type=FileType.IMAGE,
+                    transfer_method=FileTransferMethod.TOOL_FILE,
+                    related_id=value,
+                    extension=extension,
+                    mime_type=mime_type or tool_file.mimetype,
+                    filename=filename or tool_file.name,
+                    storage_key=tool_file.file_key,
+                )
+
+    logger.warning("File not found for file_ref: %s", file_ref)
+    return None

api/core/llm_generator/llm_generator.py

@@ -1,11 +1,16 @@
 import json
 import logging
 import re
-from collections.abc import Sequence
-from typing import Protocol, cast
+from collections.abc import Mapping, Sequence
+from typing import Any, Protocol, cast
 
 import json_repair
 
+from core.llm_generator.output_models import (
+    CodeNodeStructuredOutput,
+    InstructionModifyOutput,
+    SuggestedQuestionsOutput,
+)
 from core.llm_generator.output_parser.rule_config_generator import RuleConfigGeneratorOutputParser
 from core.llm_generator.output_parser.suggested_questions_after_answer import SuggestedQuestionsAfterAnswerOutputParser
 from core.llm_generator.prompts import (
@@ -393,6 +398,444 @@ class LLMGenerator:
             logger.exception("Failed to invoke LLM model, model: %s", model_config.get("name"))
             return {"output": "", "error": f"An unexpected error occurred: {str(e)}"}
 
+    @classmethod
+    def generate_with_context(
+        cls,
+        tenant_id: str,
+        workflow_id: str,
+        node_id: str,
+        parameter_name: str,
+        language: str,
+        prompt_messages: list[PromptMessage],
+        model_config: dict,
+    ) -> dict:
+        """
+        Generate extractor code node based on conversation context.
+
+        Args:
+            tenant_id: Tenant/workspace ID
+            workflow_id: Workflow ID
+            node_id: Current tool/llm node ID
+            parameter_name: Parameter name to generate code for
+            language: Code language (python3/javascript)
+            prompt_messages: Multi-turn conversation history (last message is instruction)
+            model_config: Model configuration (provider, name, completion_params)
+
+        Returns:
+            dict with CodeNodeData format:
+            - variables: Input variable selectors
+            - code_language: Code language
+            - code: Generated code
+            - outputs: Output definitions
+            - message: Explanation
+            - error: Error message if any
+        """
+        from sqlalchemy import select
+        from sqlalchemy.orm import Session
+
+        from services.workflow_service import WorkflowService
+
+        # Get workflow
+        with Session(db.engine) as session:
+            stmt = select(App).where(App.id == workflow_id)
+            app = session.scalar(stmt)
+            if not app:
+                return cls._error_response(f"App {workflow_id} not found")
+
+            workflow = WorkflowService().get_draft_workflow(app_model=app)
+            if not workflow:
+                return cls._error_response(f"Workflow for app {workflow_id} not found")
+
+        # Get upstream nodes via edge backtracking
+        upstream_nodes = cls._get_upstream_nodes(workflow.graph_dict, node_id)
+
+        # Get current node info
+        current_node = cls._get_node_by_id(workflow.graph_dict, node_id)
+        if not current_node:
+            return cls._error_response(f"Node {node_id} not found")
+
+        # Get parameter info
+        parameter_info = cls._get_parameter_info(
+            tenant_id=tenant_id,
+            node_data=current_node.get("data", {}),
+            parameter_name=parameter_name,
+        )
+
+        # Build system prompt
+        system_prompt = cls._build_extractor_system_prompt(
+            upstream_nodes=upstream_nodes,
+            current_node=current_node,
+            parameter_info=parameter_info,
+            language=language,
+        )
+
+        # Construct complete prompt_messages with system prompt
+        complete_messages: list[PromptMessage] = [
+            SystemPromptMessage(content=system_prompt),
+            *prompt_messages,
+        ]
+
+        from core.llm_generator.output_parser.structured_output import invoke_llm_with_pydantic_model
+
+        # Get model instance and schema
+        provider = model_config.get("provider", "")
+        model_name = model_config.get("name", "")
+        model_instance = ModelManager().get_model_instance(
+            tenant_id=tenant_id,
+            model_type=ModelType.LLM,
+            provider=provider,
+            model=model_name,
+        )
+
+        model_schema = model_instance.model_type_instance.get_model_schema(model_name, model_instance.credentials)
+        if not model_schema:
+            return cls._error_response(f"Model schema not found for {model_name}")
+
+        model_parameters = model_config.get("completion_params", {})
+        try:
+            response = invoke_llm_with_pydantic_model(
+                provider=provider,
+                model_schema=model_schema,
+                model_instance=model_instance,
+                prompt_messages=complete_messages,
+                output_model=CodeNodeStructuredOutput,
+                model_parameters=model_parameters,
+                stream=False,
+                tenant_id=tenant_id,
+            )
+
+            return cls._parse_code_node_output(
+                response.structured_output, language, parameter_info.get("type", "string")
+            )
+
+        except InvokeError as e:
+            return cls._error_response(str(e))
+        except Exception as e:
+            logger.exception("Failed to generate with context, model: %s", model_config.get("name"))
+            return cls._error_response(f"An unexpected error occurred: {str(e)}")
+
+    @classmethod
+    def _error_response(cls, error: str) -> dict:
+        """Return error response in CodeNodeData format."""
+        return {
+            "variables": [],
+            "code_language": "python3",
+            "code": "",
+            "outputs": {},
+            "message": "",
+            "error": error,
+        }
+
+    @classmethod
+    def generate_suggested_questions(
+        cls,
+        tenant_id: str,
+        workflow_id: str,
+        node_id: str,
+        parameter_name: str,
+        language: str,
+        model_config: dict | None = None,
+    ) -> dict:
+        """
+        Generate suggested questions for context generation.
+
+        Returns dict with questions array and error field.
+        """
+        from sqlalchemy import select
+        from sqlalchemy.orm import Session
+
+        from core.llm_generator.output_parser.structured_output import invoke_llm_with_pydantic_model
+        from services.workflow_service import WorkflowService
+
+        # Get workflow context (reuse existing logic)
+        with Session(db.engine) as session:
+            stmt = select(App).where(App.id == workflow_id)
+            app = session.scalar(stmt)
+            if not app:
+                return {"questions": [], "error": f"App {workflow_id} not found"}
+
+            workflow = WorkflowService().get_draft_workflow(app_model=app)
+            if not workflow:
+                return {"questions": [], "error": f"Workflow for app {workflow_id} not found"}
+
+        upstream_nodes = cls._get_upstream_nodes(workflow.graph_dict, node_id)
+        current_node = cls._get_node_by_id(workflow.graph_dict, node_id)
+        if not current_node:
+            return {"questions": [], "error": f"Node {node_id} not found"}
+
+        parameter_info = cls._get_parameter_info(
+            tenant_id=tenant_id,
+            node_data=current_node.get("data", {}),
+            parameter_name=parameter_name,
+        )
+
+        # Build prompt
+        system_prompt = cls._build_suggested_questions_prompt(
+            upstream_nodes=upstream_nodes,
+            current_node=current_node,
+            parameter_info=parameter_info,
+            language=language,
+        )
+
+        prompt_messages: list[PromptMessage] = [
+            SystemPromptMessage(content=system_prompt),
+        ]
+
+        # Get model instance - use default if model_config not provided
+        model_manager = ModelManager()
+        if model_config:
+            provider = model_config.get("provider", "")
+            model_name = model_config.get("name", "")
+            model_instance = model_manager.get_model_instance(
+                tenant_id=tenant_id,
+                model_type=ModelType.LLM,
+                provider=provider,
+                model=model_name,
+            )
+        else:
+            model_instance = model_manager.get_default_model_instance(
+                tenant_id=tenant_id,
+                model_type=ModelType.LLM,
+            )
+            model_name = model_instance.model
+
+        model_schema = model_instance.model_type_instance.get_model_schema(model_name, model_instance.credentials)
+        if not model_schema:
+            return {"questions": [], "error": f"Model schema not found for {model_name}"}
+
+        completion_params = model_config.get("completion_params", {}) if model_config else {}
+        model_parameters = {**completion_params, "max_tokens": 256}
+        try:
+            response = invoke_llm_with_pydantic_model(
+                provider=model_instance.provider,
+                model_schema=model_schema,
+                model_instance=model_instance,
+                prompt_messages=prompt_messages,
+                output_model=SuggestedQuestionsOutput,
+                model_parameters=model_parameters,
+                stream=False,
+                tenant_id=tenant_id,
+            )
+
+            questions = response.structured_output.get("questions", []) if response.structured_output else []
+            return {"questions": questions, "error": ""}
+
+        except InvokeError as e:
+            return {"questions": [], "error": str(e)}
+        except Exception as e:
+            logger.exception("Failed to generate suggested questions, model: %s", model_name)
+            return {"questions": [], "error": f"An unexpected error occurred: {str(e)}"}
+
+    @classmethod
+    def _build_suggested_questions_prompt(
+        cls,
+        upstream_nodes: list[dict],
+        current_node: dict,
+        parameter_info: dict,
+        language: str = "English",
+    ) -> str:
+        """Build minimal prompt for suggested questions generation."""
+        # Simplify upstream nodes to reduce tokens
+        sources = [f"{n['title']}({','.join(n.get('outputs', {}).keys())})" for n in upstream_nodes[:5]]
+        param_type = parameter_info.get("type", "string")
+        param_desc = parameter_info.get("description", "")[:100]
+
+        return f"""Suggest 3 code generation questions for extracting data.
+Sources: {", ".join(sources)}
+Target: {parameter_info.get("name")}({param_type}) - {param_desc}
+Output 3 short, practical questions in {language}."""
+
+    @classmethod
+    def _get_upstream_nodes(cls, graph_dict: Mapping[str, Any], node_id: str) -> list[dict]:
+        """
+        Get all upstream nodes via edge backtracking.
+
+        Traverses the graph backwards from node_id to collect all reachable nodes.
+        """
+        from collections import defaultdict
+
+        nodes = {n["id"]: n for n in graph_dict.get("nodes", [])}
+        edges = graph_dict.get("edges", [])
+
+        # Build reverse adjacency list
+        reverse_adj: dict[str, list[str]] = defaultdict(list)
+        for edge in edges:
+            reverse_adj[edge["target"]].append(edge["source"])
+
+        # BFS to find all upstream nodes
+        visited: set[str] = set()
+        queue = [node_id]
+        upstream: list[dict] = []
+
+        while queue:
+            current = queue.pop(0)
+            for source in reverse_adj.get(current, []):
+                if source not in visited:
+                    visited.add(source)
+                    queue.append(source)
+                    if source in nodes:
+                        upstream.append(cls._extract_node_info(nodes[source]))
+
+        return upstream
+
+    @classmethod
+    def _get_node_by_id(cls, graph_dict: Mapping[str, Any], node_id: str) -> dict | None:
+        """Get node by ID from graph."""
+        for node in graph_dict.get("nodes", []):
+            if node["id"] == node_id:
+                return node
+        return None
+
+    @classmethod
+    def _extract_node_info(cls, node: dict) -> dict:
+        """Extract minimal node info with outputs based on node type."""
+        node_type = node["data"]["type"]
+        node_data = node.get("data", {})
+
+        # Build outputs based on node type (only type, no description to reduce tokens)
+        outputs: dict[str, str] = {}
+        match node_type:
+            case "start":
+                for var in node_data.get("variables", []):
+                    name = var.get("variable", var.get("name", ""))
+                    outputs[name] = var.get("type", "string")
+            case "llm":
+                outputs["text"] = "string"
+            case "code":
+                for name, output in node_data.get("outputs", {}).items():
+                    outputs[name] = output.get("type", "string")
+            case "http-request":
+                outputs = {"body": "string", "status_code": "number", "headers": "object"}
+            case "knowledge-retrieval":
+                outputs["result"] = "array[object]"
+            case "tool":
+                outputs = {"text": "string", "json": "object"}
+            case _:
+                outputs["output"] = "string"
+
+        info: dict = {
+            "id": node["id"],
+            "title": node_data.get("title", node["id"]),
+            "outputs": outputs,
+        }
+        # Only include description if not empty
+        desc = node_data.get("desc", "")
+        if desc:
+            info["desc"] = desc
+
+        return info
+
+    @classmethod
+    def _get_parameter_info(cls, tenant_id: str, node_data: dict, parameter_name: str) -> dict:
+        """Get parameter info from tool schema using ToolManager."""
+        default_info = {"name": parameter_name, "type": "string", "description": ""}
+
+        if node_data.get("type") != "tool":
+            return default_info
+
+        try:
+            from core.app.entities.app_invoke_entities import InvokeFrom
+            from core.tools.entities.tool_entities import ToolProviderType
+            from core.tools.tool_manager import ToolManager
+
+            provider_type_str = node_data.get("provider_type", "")
+            provider_type = ToolProviderType(provider_type_str) if provider_type_str else ToolProviderType.BUILT_IN
+
+            tool_runtime = ToolManager.get_tool_runtime(
+                provider_type=provider_type,
+                provider_id=node_data.get("provider_id", ""),
+                tool_name=node_data.get("tool_name", ""),
+                tenant_id=tenant_id,
+                invoke_from=InvokeFrom.DEBUGGER,
+            )
+
+            parameters = tool_runtime.get_merged_runtime_parameters()
+            for param in parameters:
+                if param.name == parameter_name:
+                    return {
+                        "name": param.name,
+                        "type": param.type.value if hasattr(param.type, "value") else str(param.type),
+                        "description": param.llm_description
+                        or (param.human_description.en_US if param.human_description else ""),
+                        "required": param.required,
+                    }
+        except Exception as e:
+            logger.debug("Failed to get parameter info from ToolManager: %s", e)
+
+        return default_info
+
+    @classmethod
+    def _build_extractor_system_prompt(
+        cls,
+        upstream_nodes: list[dict],
+        current_node: dict,
+        parameter_info: dict,
+        language: str,
+    ) -> str:
+        """Build system prompt for extractor code generation."""
+        upstream_json = json.dumps(upstream_nodes, indent=2, ensure_ascii=False)
+        param_type = parameter_info.get("type", "string")
+        return f"""You are a code generator for workflow automation.
+
+Generate {language} code to extract/transform upstream node outputs for the target parameter.
+
+## Upstream Nodes
+{upstream_json}
+
+## Target
+Node: {current_node["data"].get("title", current_node["id"])}
+Parameter: {parameter_info.get("name")} ({param_type}) - {parameter_info.get("description", "")}
+
+## Requirements
+- Write a main function that returns type: {param_type}
+- Use value_selector format: ["node_id", "output_name"]
+"""
+
+    @classmethod
+    def _parse_code_node_output(cls, content: Mapping[str, Any] | None, language: str, parameter_type: str) -> dict:
+        """
+        Parse structured output to CodeNodeData format.
+
+        Args:
+            content: Structured output dict from invoke_llm_with_structured_output
+            language: Code language
+            parameter_type: Expected parameter type
+
+        Returns dict with variables, code_language, code, outputs, message, error.
+        """
+        if content is None:
+            return cls._error_response("Empty or invalid response from LLM")
+
+        # Validate and normalize variables
+        variables = [
+            {"variable": v.get("variable", ""), "value_selector": v.get("value_selector", [])}
+            for v in content.get("variables", [])
+            if isinstance(v, dict)
+        ]
+
+        # Convert outputs from array format [{name, type}] to dict format {name: {type}}
+        # Array format is required for OpenAI/Azure strict JSON schema compatibility
+        raw_outputs = content.get("outputs", [])
+        if isinstance(raw_outputs, list):
+            outputs = {
+                item.get("name", "result"): {"type": item.get("type", parameter_type)}
+                for item in raw_outputs
+                if isinstance(item, dict) and item.get("name")
+            }
+            if not outputs:
+                outputs = {"result": {"type": parameter_type}}
+        else:
+            outputs = raw_outputs or {"result": {"type": parameter_type}}
+
+        return {
+            "variables": variables,
+            "code_language": language,
+            "code": content.get("code", ""),
+            "outputs": outputs,
+            "message": content.get("explanation", ""),
+            "error": "",
+        }
+
     @staticmethod
     def instruction_modify_legacy(
         tenant_id: str, flow_id: str, current: str, instruction: str, model_config: dict, ideal_output: str | None
@@ -529,6 +972,10 @@ class LLMGenerator:
             provider=model_config.get("provider", ""),
             model=model_config.get("name", ""),
         )
+        model_name = model_config.get("name", "")
+        model_schema = model_instance.model_type_instance.get_model_schema(model_name, model_instance.credentials)
+        if not model_schema:
+            return {"error": f"Model schema not found for {model_name}"}
         match node_type:
             case "llm" | "agent":
                 system_prompt = LLM_MODIFY_PROMPT_SYSTEM
@@ -552,20 +999,18 @@ class LLMGenerator:
         model_parameters = {"temperature": 0.4}
 
         try:
-            response: LLMResult = model_instance.invoke_llm(
-                prompt_messages=list(prompt_messages), model_parameters=model_parameters, stream=False
-            )
+            from core.llm_generator.output_parser.structured_output import invoke_llm_with_pydantic_model
 
-            generated_raw = response.message.get_text_content()
-            first_brace = generated_raw.find("{")
-            last_brace = generated_raw.rfind("}")
-            if first_brace == -1 or last_brace == -1 or last_brace < first_brace:
-                raise ValueError(f"Could not find a valid JSON object in response: {generated_raw}")
-            json_str = generated_raw[first_brace : last_brace + 1]
-            data = json_repair.loads(json_str)
-            if not isinstance(data, dict):
-                raise TypeError(f"Expected a JSON object, but got {type(data).__name__}")
-            return data
+            response = invoke_llm_with_pydantic_model(
+                provider=model_instance.provider,
+                model_schema=model_schema,
+                model_instance=model_instance,
+                prompt_messages=list(prompt_messages),
+                output_model=InstructionModifyOutput,
+                model_parameters=model_parameters,
+                stream=False,
+            )
+            return response.structured_output or {}
         except InvokeError as e:
             error = str(e)
             return {"error": f"Failed to generate code. Error: {error}"}

api/core/llm_generator/output_models.py

@@ -0,0 +1,67 @@
+from __future__ import annotations
+
+from pydantic import BaseModel, ConfigDict, Field
+
+from core.variables.types import SegmentType
+
+
+class SuggestedQuestionsOutput(BaseModel):
+    """Output model for suggested questions generation."""
+
+    model_config = ConfigDict(extra="forbid")
+
+    questions: list[str] = Field(
+        min_length=3,
+        max_length=3,
+        description="Exactly 3 suggested follow-up questions for the user",
+    )
+
+
+class VariableSelectorOutput(BaseModel):
+    """Variable selector mapping code variable to upstream node output.
+
+    Note: Separate from VariableSelector to ensure 'additionalProperties: false'
+    in JSON schema for OpenAI/Azure strict mode.
+    """
+
+    model_config = ConfigDict(extra="forbid")
+
+    variable: str = Field(description="Variable name used in the generated code")
+    value_selector: list[str] = Field(description="Path to upstream node output, format: [node_id, output_name]")
+
+
+class CodeNodeOutputItem(BaseModel):
+    """Single output variable definition.
+
+    Note: OpenAI/Azure strict mode requires 'additionalProperties: false' and
+    does not support dynamic object keys, so outputs use array format.
+    """
+
+    model_config = ConfigDict(extra="forbid")
+
+    name: str = Field(description="Output variable name returned by the main function")
+    type: SegmentType = Field(description="Data type of the output variable")
+
+
+class CodeNodeStructuredOutput(BaseModel):
+    """Structured output for code node generation."""
+
+    model_config = ConfigDict(extra="forbid")
+
+    variables: list[VariableSelectorOutput] = Field(
+        description="Input variables mapping code variables to upstream node outputs"
+    )
+    code: str = Field(description="Generated code with a main function that processes inputs and returns outputs")
+    outputs: list[CodeNodeOutputItem] = Field(
+        description="Output variable definitions specifying name and type for each return value"
+    )
+    explanation: str = Field(description="Brief explanation of what the generated code does")
+
+
+class InstructionModifyOutput(BaseModel):
+    """Output model for instruction-based prompt modification."""
+
+    model_config = ConfigDict(extra="forbid")
+
+    modified: str = Field(description="The modified prompt content after applying the instruction")
+    message: str = Field(description="Brief explanation of what changes were made")

api/core/llm_generator/output_parser/file_ref.py

@@ -0,0 +1,188 @@
+"""
+File reference detection and conversion for structured output.
+
+This module provides utilities to:
+1. Detect file reference fields in JSON Schema (format: "dify-file-ref")
+2. Convert file ID strings to File objects after LLM returns
+"""
+
+import uuid
+from collections.abc import Mapping
+from typing import Any
+
+from core.file import File
+from core.variables.segments import ArrayFileSegment, FileSegment
+from factories.file_factory import build_from_mapping
+
+FILE_REF_FORMAT = "dify-file-ref"
+
+
+def is_file_ref_property(schema: dict) -> bool:
+    """Check if a schema property is a file reference."""
+    return schema.get("type") == "string" and schema.get("format") == FILE_REF_FORMAT
+
+
+def detect_file_ref_fields(schema: Mapping[str, Any], path: str = "") -> list[str]:
+    """
+    Recursively detect file reference fields in schema.
+
+    Args:
+        schema: JSON Schema to analyze
+        path: Current path in the schema (used for recursion)
+
+    Returns:
+        List of JSON paths containing file refs, e.g., ["image_id", "files[*]"]
+    """
+    file_ref_paths: list[str] = []
+    schema_type = schema.get("type")
+
+    if schema_type == "object":
+        for prop_name, prop_schema in schema.get("properties", {}).items():
+            current_path = f"{path}.{prop_name}" if path else prop_name
+
+            if is_file_ref_property(prop_schema):
+                file_ref_paths.append(current_path)
+            elif isinstance(prop_schema, dict):
+                file_ref_paths.extend(detect_file_ref_fields(prop_schema, current_path))
+
+    elif schema_type == "array":
+        items_schema = schema.get("items", {})
+        array_path = f"{path}[*]" if path else "[*]"
+
+        if is_file_ref_property(items_schema):
+            file_ref_paths.append(array_path)
+        elif isinstance(items_schema, dict):
+            file_ref_paths.extend(detect_file_ref_fields(items_schema, array_path))
+
+    return file_ref_paths
+
+
+def convert_file_refs_in_output(
+    output: Mapping[str, Any],
+    json_schema: Mapping[str, Any],
+    tenant_id: str,
+) -> dict[str, Any]:
+    """
+    Convert file ID strings to File objects based on schema.
+
+    Args:
+        output: The structured_output from LLM result
+        json_schema: The original JSON schema (to detect file ref fields)
+        tenant_id: Tenant ID for file lookup
+
+    Returns:
+        Output with file references converted to File objects
+    """
+    file_ref_paths = detect_file_ref_fields(json_schema)
+    if not file_ref_paths:
+        return dict(output)
+
+    result = _deep_copy_dict(output)
+
+    for path in file_ref_paths:
+        _convert_path_in_place(result, path.split("."), tenant_id)
+
+    return result
+
+
+def _deep_copy_dict(obj: Mapping[str, Any]) -> dict[str, Any]:
+    """Deep copy a mapping to a mutable dict."""
+    result: dict[str, Any] = {}
+    for key, value in obj.items():
+        if isinstance(value, Mapping):
+            result[key] = _deep_copy_dict(value)
+        elif isinstance(value, list):
+            result[key] = [_deep_copy_dict(item) if isinstance(item, Mapping) else item for item in value]
+        else:
+            result[key] = value
+    return result
+
+
+def _convert_path_in_place(obj: dict, path_parts: list[str], tenant_id: str) -> None:
+    """Convert file refs at the given path in place, wrapping in Segment types."""
+    if not path_parts:
+        return
+
+    current = path_parts[0]
+    remaining = path_parts[1:]
+
+    # Handle array notation like "files[*]"
+    if current.endswith("[*]"):
+        key = current[:-3] if current != "[*]" else None
+        target = obj.get(key) if key else obj
+
+        if isinstance(target, list):
+            if remaining:
+                # Nested array with remaining path - recurse into each item
+                for item in target:
+                    if isinstance(item, dict):
+                        _convert_path_in_place(item, remaining, tenant_id)
+            else:
+                # Array of file IDs - convert all and wrap in ArrayFileSegment
+                files: list[File] = []
+                for item in target:
+                    file = _convert_file_id(item, tenant_id)
+                    if file is not None:
+                        files.append(file)
+                # Replace the array with ArrayFileSegment
+                if key:
+                    obj[key] = ArrayFileSegment(value=files)
+        return
+
+    if not remaining:
+        # Leaf node - convert the value and wrap in FileSegment
+        if current in obj:
+            file = _convert_file_id(obj[current], tenant_id)
+            if file is not None:
+                obj[current] = FileSegment(value=file)
+            else:
+                obj[current] = None
+    else:
+        # Recurse into nested object
+        if current in obj and isinstance(obj[current], dict):
+            _convert_path_in_place(obj[current], remaining, tenant_id)
+
+
+def _convert_file_id(file_id: Any, tenant_id: str) -> File | None:
+    """
+    Convert a file ID string to a File object.
+
+    Tries multiple file sources in order:
+    1. ToolFile (files generated by tools/workflows)
+    2. UploadFile (files uploaded by users)
+    """
+    if not isinstance(file_id, str):
+        return None
+
+    # Validate UUID format
+    try:
+        uuid.UUID(file_id)
+    except ValueError:
+        return None
+
+    # Try ToolFile first (files generated by tools/workflows)
+    try:
+        return build_from_mapping(
+            mapping={
+                "transfer_method": "tool_file",
+                "tool_file_id": file_id,
+            },
+            tenant_id=tenant_id,
+        )
+    except ValueError:
+        pass
+
+    # Try UploadFile (files uploaded by users)
+    try:
+        return build_from_mapping(
+            mapping={
+                "transfer_method": "local_file",
+                "upload_file_id": file_id,
+            },
+            tenant_id=tenant_id,
+        )
+    except ValueError:
+        pass
+
+    # File not found in any source
+    return None

api/core/llm_generator/output_parser/structured_output.py

@@ -2,12 +2,13 @@ import json
 from collections.abc import Generator, Mapping, Sequence
 from copy import deepcopy
 from enum import StrEnum
-from typing import Any, Literal, cast, overload
+from typing import Any, Literal, TypeVar, cast, overload
 
 import json_repair
-from pydantic import TypeAdapter, ValidationError
+from pydantic import BaseModel, TypeAdapter, ValidationError
 
 from core.llm_generator.output_parser.errors import OutputParserError
+from core.llm_generator.output_parser.file_ref import convert_file_refs_in_output
 from core.llm_generator.prompts import STRUCTURED_OUTPUT_PROMPT
 from core.model_manager import ModelInstance
 from core.model_runtime.callbacks.base_callback import Callback
@@ -43,6 +44,9 @@ class SpecialModelType(StrEnum):
     OLLAMA = "ollama"
 
 
+T = TypeVar("T", bound=BaseModel)
+
+
 @overload
 def invoke_llm_with_structured_output(
     *,
@@ -57,6 +61,7 @@ def invoke_llm_with_structured_output(
     stream: Literal[True],
     user: str | None = None,
     callbacks: list[Callback] | None = None,
+    tenant_id: str | None = None,
 ) -> Generator[LLMResultChunkWithStructuredOutput, None, None]: ...
 @overload
 def invoke_llm_with_structured_output(
@@ -72,6 +77,7 @@ def invoke_llm_with_structured_output(
     stream: Literal[False],
     user: str | None = None,
     callbacks: list[Callback] | None = None,
+    tenant_id: str | None = None,
 ) -> LLMResultWithStructuredOutput: ...
 @overload
 def invoke_llm_with_structured_output(
@@ -87,6 +93,7 @@ def invoke_llm_with_structured_output(
     stream: bool = True,
     user: str | None = None,
     callbacks: list[Callback] | None = None,
+    tenant_id: str | None = None,
 ) -> LLMResultWithStructuredOutput | Generator[LLMResultChunkWithStructuredOutput, None, None]: ...
 def invoke_llm_with_structured_output(
     *,
@@ -101,23 +108,30 @@ def invoke_llm_with_structured_output(
     stream: bool = True,
     user: str | None = None,
     callbacks: list[Callback] | None = None,
+    tenant_id: str | None = None,
 ) -> LLMResultWithStructuredOutput | Generator[LLMResultChunkWithStructuredOutput, None, None]:
     """
-    Invoke large language model with structured output
-    1. This method invokes model_instance.invoke_llm with json_schema
-    2. Try to parse the result as structured output
+    Invoke large language model with structured output.
 
+    This method invokes model_instance.invoke_llm with json_schema and parses
+    the result as structured output.
+
+    :param provider: model provider name
+    :param model_schema: model schema entity
+    :param model_instance: model instance to invoke
     :param prompt_messages: prompt messages
-    :param json_schema: json schema
+    :param json_schema: json schema for structured output
     :param model_parameters: model parameters
     :param tools: tools for tool calling
     :param stop: stop words
     :param stream: is stream response
     :param user: unique user id
     :param callbacks: callbacks
+    :param tenant_id: tenant ID for file reference conversion. When provided and
+                      json_schema contains file reference fields (format: "dify-file-ref"),
+                      file IDs in the output will be automatically converted to File objects.
     :return: full response or stream response chunk generator result
     """
-
     # handle native json schema
     model_parameters_with_json_schema: dict[str, Any] = {
         **(model_parameters or {}),
@@ -153,8 +167,18 @@ def invoke_llm_with_structured_output(
                 f"Failed to parse structured output, LLM result is not a string: {llm_result.message.content}"
             )
 
+        structured_output = _parse_structured_output(llm_result.message.content)
+
+        # Convert file references if tenant_id is provided
+        if tenant_id is not None:
+            structured_output = convert_file_refs_in_output(
+                output=structured_output,
+                json_schema=json_schema,
+                tenant_id=tenant_id,
+            )
+
         return LLMResultWithStructuredOutput(
-            structured_output=_parse_structured_output(llm_result.message.content),
+            structured_output=structured_output,
             model=llm_result.model,
             message=llm_result.message,
             usage=llm_result.usage,
@@ -186,8 +210,18 @@ def invoke_llm_with_structured_output(
                     delta=event.delta,
                 )
 
+            structured_output = _parse_structured_output(result_text)
+
+            # Convert file references if tenant_id is provided
+            if tenant_id is not None:
+                structured_output = convert_file_refs_in_output(
+                    output=structured_output,
+                    json_schema=json_schema,
+                    tenant_id=tenant_id,
+                )
+
             yield LLMResultChunkWithStructuredOutput(
-                structured_output=_parse_structured_output(result_text),
+                structured_output=structured_output,
                 model=model_schema.model,
                 prompt_messages=prompt_messages,
                 system_fingerprint=system_fingerprint,
@@ -202,6 +236,87 @@ def invoke_llm_with_structured_output(
         return generator()
 
 
+@overload
+def invoke_llm_with_pydantic_model(
+    *,
+    provider: str,
+    model_schema: AIModelEntity,
+    model_instance: ModelInstance,
+    prompt_messages: Sequence[PromptMessage],
+    output_model: type[T],
+    model_parameters: Mapping | None = None,
+    tools: Sequence[PromptMessageTool] | None = None,
+    stop: list[str] | None = None,
+    stream: Literal[False] = False,
+    user: str | None = None,
+    callbacks: list[Callback] | None = None,
+    tenant_id: str | None = None,
+) -> LLMResultWithStructuredOutput: ...
+
+
+def invoke_llm_with_pydantic_model(
+    *,
+    provider: str,
+    model_schema: AIModelEntity,
+    model_instance: ModelInstance,
+    prompt_messages: Sequence[PromptMessage],
+    output_model: type[T],
+    model_parameters: Mapping | None = None,
+    tools: Sequence[PromptMessageTool] | None = None,
+    stop: list[str] | None = None,
+    stream: bool = False,
+    user: str | None = None,
+    callbacks: list[Callback] | None = None,
+    tenant_id: str | None = None,
+) -> LLMResultWithStructuredOutput:
+    """
+    Invoke large language model with a Pydantic output model.
+
+    This helper generates a JSON schema from the Pydantic model, invokes the
+    structured-output LLM path, and validates the result in non-streaming mode.
+    """
+    if stream:
+        raise ValueError("invoke_llm_with_pydantic_model only supports stream=False")
+
+    json_schema = _schema_from_pydantic(output_model)
+    result = invoke_llm_with_structured_output(
+        provider=provider,
+        model_schema=model_schema,
+        model_instance=model_instance,
+        prompt_messages=prompt_messages,
+        json_schema=json_schema,
+        model_parameters=model_parameters,
+        tools=tools,
+        stop=stop,
+        stream=False,
+        user=user,
+        callbacks=callbacks,
+        tenant_id=tenant_id,
+    )
+
+    structured_output = result.structured_output
+    if structured_output is None:
+        raise OutputParserError("Structured output is empty")
+
+    validated_output = _validate_structured_output(output_model, structured_output)
+    return result.model_copy(update={"structured_output": validated_output})
+
+
+def _schema_from_pydantic(output_model: type[BaseModel]) -> dict[str, Any]:
+    return output_model.model_json_schema()
+
+
+def _validate_structured_output(
+    output_model: type[T],
+    structured_output: Mapping[str, Any],
+) -> dict[str, Any]:
+    try:
+        validated_output = output_model.model_validate(structured_output)
+    except ValidationError as exc:
+        raise OutputParserError(f"Structured output validation failed: {exc}") from exc
+    return validated_output.model_dump(mode="python")
+
+
 def _handle_native_json_schema(
     provider: str,
     model_schema: AIModelEntity,

api/core/llm_generator/utils.py

@@ -0,0 +1,45 @@
+"""Utility functions for LLM generator."""
+
+from core.model_runtime.entities.message_entities import (
+    AssistantPromptMessage,
+    PromptMessage,
+    PromptMessageRole,
+    SystemPromptMessage,
+    ToolPromptMessage,
+    UserPromptMessage,
+)
+
+
+def deserialize_prompt_messages(messages: list[dict]) -> list[PromptMessage]:
+    """
+    Deserialize list of dicts to list[PromptMessage].
+
+    Expected format:
+        [
+            {"role": "user", "content": "..."},
+            {"role": "assistant", "content": "..."},
+        ]
+    """
+    result: list[PromptMessage] = []
+    for msg in messages:
+        role = PromptMessageRole.value_of(msg["role"])
+        content = msg.get("content", "")
+
+        match role:
+            case PromptMessageRole.USER:
+                result.append(UserPromptMessage(content=content))
+            case PromptMessageRole.ASSISTANT:
+                result.append(AssistantPromptMessage(content=content))
+            case PromptMessageRole.SYSTEM:
+                result.append(SystemPromptMessage(content=content))
+            case PromptMessageRole.TOOL:
+                result.append(ToolPromptMessage(content=content, tool_call_id=msg.get("tool_call_id", "")))
+
+    return result
+
+
+def serialize_prompt_messages(messages: list[PromptMessage]) -> list[dict]:
+    """
+    Serialize list[PromptMessage] to list of dicts.
+    """
+    return [{"role": msg.role.value, "content": msg.content} for msg in messages]