From 393d8c7c37591b0b6ff1f5fa6af7db9beb99e9e5 Mon Sep 17 00:00:00 2001 From: zhuchao Date: Tue, 24 Feb 2026 04:38:05 +0000 Subject: [PATCH 1/7] chore: add .worktrees/ to .gitignore Co-Authored-By: Claude Opus 4.6 --- .gitignore | 3 +++ 1 file changed, 3 insertions(+) diff --git a/.gitignore b/.gitignore index e300b2e..bbf8dae 100644 --- a/.gitignore +++ b/.gitignore @@ -8,4 +8,7 @@ projects/queue_data worktree .idea/* +# Git worktrees +.worktrees/ + .idea/ From 0a6ec955185ccba3fc7ee842bb0a33ceff16d3ca Mon Sep 17 00:00:00 2001 From: shuirong Date: Mon, 16 Mar 2026 16:18:50 +0900 Subject: [PATCH 2/7] =?UTF-8?q?feat(CI):=20=E6=B7=BB=E5=8A=A0=20onprem-dev?= =?UTF-8?q?=20=E7=8E=AF=E5=A2=83=E7=9A=84=E6=9E=84=E5=BB=BA=E5=92=8C?= =?UTF-8?q?=E9=83=A8=E7=BD=B2=E9=85=8D=E7=BD=AE?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit 在 CircleCI 配置中新增 onprem-dev 环境的 build-and-push 和 deploy 任务，部署到 cluster-for-B 的 onprem-dev 命名空间 Co-Authored-By: Claude Opus 4.6 (1M context) --- .circleci/config.yml | 27 +++++++++++++++++++++++++++ 1 file changed, 27 insertions(+) diff --git a/.circleci/config.yml b/.circleci/config.yml index b0a1532..d0b7173 100644 --- a/.circleci/config.yml +++ b/.circleci/config.yml @@ -193,3 +193,30 @@ workflows: branches: only: - onprem + # 为 onprem-dev 环境部署 + - build-and-push: + name: build-for-onprem-dev + context: + - ecr-new + path: . + dockerfile: Dockerfile + repo: catalog-agent + docker-tag: '' + filters: + branches: + only: + - onprem + - deploy: + name: deploy-for-onprem-dev + docker-tag: '' + path: '/home/ubuntu/cluster-for-B/onprem-dev/catalog-agent/deploy.yaml' + deploy-name: catalog-agent + deploy-namespace: onprem-dev + context: + - ecr-new + filters: + branches: + only: + - onprem + requires: + - build-for-onprem-dev From deb78a76253a3a3855965858443e92b9f0b2da8a Mon Sep 17 00:00:00 2001 From: autobee-sparticle Date: Tue, 17 Mar 2026 10:37:49 +0900 Subject: [PATCH 3/7] fix: improve memory extraction for colloquial/informal speech (#16) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit * chore: add .worktrees/ to .gitignore Co-Authored-By: Claude Opus 4.6 * feat(CI): 添加 onprem-dev 环境的构建和部署配置 在 CircleCI 配置中新增 onprem-dev 环境的 build-and-push 和 deploy 任务，部署到 cluster-for-B 的 onprem-dev 命名空间 Co-Authored-By: Claude Opus 4.6 (1M context) * fix: improve memory extraction for colloquial/informal speech Add semantic completeness rules and multilingual few-shot examples to FACT_RETRIEVAL_PROMPT to prevent truncated or semantically incorrect memory extraction. Specifically addresses Japanese casual speech where particles (が, を, に) are often omitted. Closes sparticleinc/mygpt-frontend#2125 Co-Authored-By: Claude Opus 4.6 --------- Co-authored-by: zhuchao Co-authored-by: Claude Opus 4.6 Co-authored-by: shuirong --- .circleci/config.yml | 27 +++++++++++++++++++++++++++ prompt/FACT_RETRIEVAL_PROMPT.md | 20 ++++++++++++++++++++ 2 files changed, 47 insertions(+) diff --git a/.circleci/config.yml b/.circleci/config.yml index b0a1532..d0b7173 100644 --- a/.circleci/config.yml +++ b/.circleci/config.yml @@ -193,3 +193,30 @@ workflows: branches: only: - onprem + # 为 onprem-dev 环境部署 + - build-and-push: + name: build-for-onprem-dev + context: + - ecr-new + path: . + dockerfile: Dockerfile + repo: catalog-agent + docker-tag: '' + filters: + branches: + only: + - onprem + - deploy: + name: deploy-for-onprem-dev + docker-tag: '' + path: '/home/ubuntu/cluster-for-B/onprem-dev/catalog-agent/deploy.yaml' + deploy-name: catalog-agent + deploy-namespace: onprem-dev + context: + - ecr-new + filters: + branches: + only: + - onprem + requires: + - build-for-onprem-dev diff --git a/prompt/FACT_RETRIEVAL_PROMPT.md b/prompt/FACT_RETRIEVAL_PROMPT.md index 27777e6..1c2f93e 100644 --- a/prompt/FACT_RETRIEVAL_PROMPT.md +++ b/prompt/FACT_RETRIEVAL_PROMPT.md @@ -83,6 +83,21 @@ Output: {{"facts" : ["Mike Smith helped with bug fix", "Contact: Mike Smith (col Input: Mike is coming to the meeting tomorrow. Output: {{"facts" : ["Mike Smith is coming to the meeting tomorrow", "Contact: Mike Smith (colleague, also referred as Mike) - DEFAULT when user says 'Mike'"]}} +Input: 私は林檎好きです +Output: {{"facts" : ["林檎が好き"]}} + +Input: コーヒー飲みたい、毎朝 +Output: {{"facts" : ["毎朝コーヒーを飲みたい"]}} + +Input: 昨日映画見た、すごくよかった +Output: {{"facts" : ["昨日映画を見た", "映画がすごくよかった"]}} + +Input: 我喜欢吃苹果 +Output: {{"facts" : ["喜欢吃苹果"]}} + +Input: 나는 사과를 좋아해 +Output: {{"facts" : ["사과를 좋아함"]}} + Return the facts and preferences in a json format as shown above. Remember the following: @@ -93,6 +108,11 @@ Remember the following: - If you do not find anything relevant in the below conversation, you can return an empty list corresponding to the "facts" key. - Create the facts based on the user and assistant messages only. Do not pick anything from the system messages. - Make sure to return the response in the format mentioned in the examples. The response should be in json with a key as "facts" and corresponding value will be a list of strings. +- **CRITICAL for Semantic Completeness**: + - Each extracted fact MUST preserve the complete semantic meaning. Never truncate or drop key parts of the meaning. + - For colloquial or grammatically informal expressions (common in spoken Japanese, Chinese, Korean, etc.), understand the full intended meaning and record it in a clear, semantically complete form. + - In Japanese, spoken language often omits particles (e.g., が, を, に). When extracting facts, include the necessary particles to make the meaning unambiguous. For example: "私は林檎好きです" should be understood as "林檎が好き" (likes apples), not literally "私は林檎好き". + - When the user expresses a preference or opinion in casual speech, record the core preference/opinion clearly. Remove the subject pronoun (私は/I) since facts are about the user by default, but keep all other semantic components intact. - **CRITICAL for Contact/Relationship Tracking**: - ALWAYS use the "Contact: [name] (relationship/context)" format when recording people - When you see a short name that matches a known full name, record as "Contact: [Full Name] (relationship, also referred as [Short Name])" From a161e43421bcd6e66e725525a02300bd94153569 Mon Sep 17 00:00:00 2001 From: autobee-sparticle Date: Tue, 17 Mar 2026 11:14:02 +0900 Subject: [PATCH 4/7] feat: add POST /api/v1/memory endpoint for realtime conversation memory (#17) * feat: add POST /api/v1/memory endpoint for realtime conversation memory Add memory extraction API that accepts conversation messages and stores them via Mem0. This enables realtime voice sessions to save memories through the same pipeline as chat conversations. Fixes: sparticleinc/mygpt-frontend#2126 Co-Authored-By: Claude Opus 4.6 * fix: address code review findings for memory API - Use Literal["user","assistant"] for role field validation - Add Field constraints (min_length, max_length=200) - Track and report pairs_failed in response - Hide internal exception details from HTTP response - Remove unused authorization parameter (internal API) Co-Authored-By: Claude Opus 4.6 --------- Co-authored-by: zhuchao Co-authored-by: Claude Opus 4.6 --- routes/memory.py | 112 +++++++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 109 insertions(+), 3 deletions(-) diff --git a/routes/memory.py b/routes/memory.py index a0b6a95..c29ae53 100644 --- a/routes/memory.py +++ b/routes/memory.py @@ -1,13 +1,13 @@ """ Memory 管理 API 路由 -提供记忆查看和删除功能 +提供记忆查看、添加和删除功能 """ import logging -from typing import Optional, List, Dict, Any +from typing import Literal, Optional, List, Dict, Any from fastapi import APIRouter, HTTPException, Header, Query from fastapi.responses import JSONResponse -from pydantic import BaseModel +from pydantic import BaseModel, Field logger = logging.getLogger('app') @@ -33,6 +33,26 @@ class DeleteAllResponse(BaseModel): deleted_count: int +class ConversationMessage(BaseModel): + """对话消息""" + role: Literal["user", "assistant"] + content: str = Field(..., min_length=1) + + +class AddMemoryRequest(BaseModel): + """添加记忆的请求体""" + bot_id: str = Field(..., min_length=1) + user_id: str = Field(..., min_length=1) + messages: List[ConversationMessage] = Field(..., max_length=200) + + +class AddMemoryResponse(BaseModel): + """添加记忆的响应""" + success: bool + pairs_processed: int + pairs_failed: int = 0 + + async def get_user_identifier_from_request( authorization: Optional[str], user_id: Optional[str] = None @@ -63,6 +83,92 @@ async def get_user_identifier_from_request( ) +@router.post("/memory", response_model=AddMemoryResponse) +async def add_memory_from_conversation(data: AddMemoryRequest): + """ + 从对话消息中提取并保存记忆 + + 将用户和助手的对话配对，通过 Mem0 提取关键事实并存储。 + 用于 realtime 语音对话等不经过 Agent 中间件的场景。 + 此端点供内部服务调用（如 felo-mygpt），不暴露给外部用户。 + """ + try: + from agent.mem0_manager import get_mem0_manager + from utils.settings import MEM0_ENABLED + + if not MEM0_ENABLED: + raise HTTPException( + status_code=503, + detail="Memory feature is not enabled" + ) + + if not data.messages: + return AddMemoryResponse(success=True, pairs_processed=0) + + manager = get_mem0_manager() + + # 将消息配对为 user-assistant 对，然后调用 add_memory + pairs_processed = 0 + pairs_failed = 0 + i = 0 + while i < len(data.messages): + msg = data.messages[i] + if msg.role == 'user': + # 收集连续的 user 消息 + user_contents = [msg.content] + j = i + 1 + while j < len(data.messages) and data.messages[j].role == 'user': + user_contents.append(data.messages[j].content) + j += 1 + + user_text = '\n'.join(user_contents) + + # 检查是否有对应的 assistant 回复 + assistant_text = "" + if j < len(data.messages) and data.messages[j].role == 'assistant': + assistant_text = data.messages[j].content or "" + j += 1 + + if user_text and assistant_text: + conversation_text = f"User: {user_text}\nAssistant: {assistant_text}" + try: + await manager.add_memory( + text=conversation_text, + user_id=data.user_id, + agent_id=data.bot_id, + metadata={"type": "realtime_conversation"}, + ) + pairs_processed += 1 + except Exception as pair_error: + pairs_failed += 1 + logger.error( + f"Failed to add memory for pair: {pair_error}" + ) + + i = j + else: + i += 1 + + logger.info( + f"Added {pairs_processed} memory pairs (failed={pairs_failed}) " + f"for user={data.user_id}, bot={data.bot_id}" + ) + return AddMemoryResponse( + success=pairs_failed == 0, + pairs_processed=pairs_processed, + pairs_failed=pairs_failed, + ) + + except HTTPException: + raise + except Exception as e: + logger.error(f"Failed to add memory from conversation: {e}") + raise HTTPException( + status_code=500, + detail="Failed to add memory from conversation" + ) + + @router.get("/memory", response_model=MemoryListResponse) async def get_memories( bot_id: str = Query(..., description="Bot ID (对应 agent_id)"), From 822589eab663e183c34c492d1ebc28ad53187dda Mon Sep 17 00:00:00 2001 From: shuirong Date: Wed, 18 Mar 2026 16:16:53 +0900 Subject: [PATCH 5/7] =?UTF-8?q?ci:=20=E9=87=8D=E6=96=B0=E8=B0=83=E6=95=B4C?= =?UTF-8?q?I?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- .circleci/config.yml | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/.circleci/config.yml b/.circleci/config.yml index d0b7173..0d541a6 100644 --- a/.circleci/config.yml +++ b/.circleci/config.yml @@ -108,7 +108,7 @@ workflows: filters: branches: only: - - master + - dev - deploy: name: deploy-for-test docker-tag: '' @@ -120,7 +120,7 @@ workflows: filters: branches: only: - - master + - dev requires: - build-for-test - build-and-push: @@ -183,7 +183,7 @@ workflows: filters: branches: only: - - onprem + - onprem-release - docker-hub-build-push: name: docker-hub-build-push-arm repo: gptbasesparticle/catalog-agent @@ -192,7 +192,7 @@ workflows: filters: branches: only: - - onprem + - onprem-release # 为 onprem-dev 环境部署 - build-and-push: name: build-for-onprem-dev @@ -205,7 +205,7 @@ workflows: filters: branches: only: - - onprem + - onprem-dev - deploy: name: deploy-for-onprem-dev docker-tag: '' @@ -217,6 +217,6 @@ workflows: filters: branches: only: - - onprem + - onprem-dev requires: - build-for-onprem-dev From 132147120277726fa80bf4af7c345ab524b82822 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=E6=9C=B1=E6=BD=AE?= Date: Fri, 3 Apr 2026 19:57:34 +0800 Subject: [PATCH 6/7] add file_manage doc --- docs/file_manage_apis.md | 435 +++++++++++++++++++++++++++++++++++++++ 1 file changed, 435 insertions(+) create mode 100644 docs/file_manage_apis.md diff --git a/docs/file_manage_apis.md b/docs/file_manage_apis.md new file mode 100644 index 0000000..6d9f612 --- /dev/null +++ b/docs/file_manage_apis.md @@ -0,0 +1,435 @@ +# 文件管理 API 接口文档 + +**Base URL:** `/api/v1/file-manager` + +--- + +## 1. 列出目录内容 + +**GET** `/api/v1/file-manager/list` + +| 参数 | 类型 | 位置 | 必填 | 说明 | +|------|------|------|------|------| +| path | string | query | 否 | 相对路径，空字符串返回根目录（projects/prompt） | +| recursive | bool | query | 否 | 是否递归列出子目录，默认 false | + +```bash +curl -X GET "/api/v1/file-manager/list?path=projects&recursive=false" +``` + +响应示例： +```json +{ + "success": true, + "path": "projects", + "items": [ + { + "name": "report.pdf", + "path": "projects/report.pdf", + "type": "file", + "size": 102400, + "modified": 1700000000.0, + "created": 1699000000.0 + } + ], + "total": 1 +} +``` + +--- + +## 2. 上传文件 + +**POST** `/api/v1/file-manager/upload` + +请求方式：表单数据上传（multipart/form-data） + +| 参数 | 类型 | 位置 | 必填 | 说明 | +|------|------|------|------|------| +| file | file | form | 是 | 上传的文件 | +| path | string | form | 否 | 目标目录路径，默认 `projects` | + +```bash +curl -X POST "/api/v1/file-manager/upload" \ + -F "file=@document.pdf" \ + -F "path=projects/dataset" +``` + +响应示例： +```json +{ + "success": true, + "message": "文件上传成功", + "filename": "document.pdf", + "original_filename": "document.pdf", + "path": "projects/dataset/document.pdf", + "size": 102400 +} +``` + +> 注：文件名中的空格会自动替换为下划线。 + +--- + +## 3. 下载文件 + +**GET** `/api/v1/file-manager/download/{file_path}` + +| 参数 | 类型 | 位置 | 必填 | 说明 | +|------|------|------|------|------| +| file_path | string | path | 是 | 文件相对路径 | + +```bash +curl -X GET "/api/v1/file-manager/download/projects/report.pdf" -o report.pdf +``` + +响应：文件二进制流，`Content-Type` 根据文件类型自动推断。 + +--- + +## 4. 删除文件或目录 + +**DELETE** `/api/v1/file-manager/delete` + +| 参数 | 类型 | 位置 | 必填 | 说明 | +|------|------|------|------|------| +| path | string | query | 是 | 要删除的路径 | + +```bash +curl -X DELETE "/api/v1/file-manager/delete?path=projects/report.pdf" +``` + +响应示例： +```json +{ + "success": true, + "message": "文件删除成功", + "path": "projects/report.pdf" +} +``` + +--- + +## 5. 创建文件夹 + +**POST** `/api/v1/file-manager/create-folder` + +请求方式：JSON + +| 参数 | 类型 | 位置 | 必填 | 说明 | +|------|------|------|------|------| +| path | string | body | 否 | 父目录路径，默认 `projects` | +| name | string | body | 是 | 文件夹名称 | + +```bash +curl -X POST "/api/v1/file-manager/create-folder" \ + -H "Content-Type: application/json" \ + -d '{"path": "projects", "name": "new-folder"}' +``` + +响应示例： +```json +{ + "success": true, + "message": "文件夹创建成功", + "path": "projects/new-folder" +} +``` + +--- + +## 6. 重命名文件或文件夹 + +**POST** `/api/v1/file-manager/rename` + +| 参数 | 类型 | 位置 | 必填 | 说明 | +|------|------|------|------|------| +| old_path | string | query | 是 | 原路径 | +| new_name | string | query | 是 | 新名称 | + +```bash +curl -X POST "/api/v1/file-manager/rename?old_path=projects/old-name.pdf&new_name=new-name.pdf" +``` + +响应示例： +```json +{ + "success": true, + "message": "重命名成功", + "old_path": "projects/old-name.pdf", + "new_path": "projects/new-name.pdf" +} +``` + +--- + +## 7. 移动文件或文件夹 + +**POST** `/api/v1/file-manager/move` + +| 参数 | 类型 | 位置 | 必填 | 说明 | +|------|------|------|------|------| +| source_path | string | query | 是 | 源路径 | +| target_path | string | query | 是 | 目标路径 | + +```bash +curl -X POST "/api/v1/file-manager/move?source_path=projects/a.pdf&target_path=projects/sub/a.pdf" +``` + +响应示例： +```json +{ + "success": true, + "message": "移动成功", + "source_path": "projects/a.pdf", + "target_path": "projects/sub/a.pdf" +} +``` + +--- + +## 8. 复制文件或文件夹 + +**POST** `/api/v1/file-manager/copy` + +| 参数 | 类型 | 位置 | 必填 | 说明 | +|------|------|------|------|------| +| source_path | string | query | 是 | 源路径 | +| target_path | string | query | 是 | 目标路径 | + +```bash +curl -X POST "/api/v1/file-manager/copy?source_path=projects/a.pdf&target_path=projects/a-copy.pdf" +``` + +响应示例： +```json +{ + "success": true, + "message": "复制成功", + "source_path": "projects/a.pdf", + "target_path": "projects/a-copy.pdf" +} +``` + +--- + +## 9. 搜索文件 + +**GET** `/api/v1/file-manager/search` + +| 参数 | 类型 | 位置 | 必填 | 说明 | +|------|------|------|------|------| +| query | string | query | 是 | 搜索关键词（匹配文件名） | +| path | string | query | 否 | 搜索路径，为空则搜索所有支持目录 | +| file_type | string | query | 否 | 文件类型过滤（如 `.pdf`） | + +```bash +curl -X GET "/api/v1/file-manager/search?query=report&path=projects&file_type=.pdf" +``` + +响应示例： +```json +{ + "success": true, + "query": "report", + "path": "projects", + "results": [ + { + "name": "report.pdf", + "path": "projects/report.pdf", + "type": "file" + } + ], + "total": 1 +} +``` + +--- + +## 10. 读取文件内容 + +**GET** `/api/v1/file-manager/read` + +| 参数 | 类型 | 位置 | 必填 | 说明 | +|------|------|------|------|------| +| path | string | query | 是 | 文件相对路径 | + +```bash +curl -X GET "/api/v1/file-manager/read?path=projects/config.json" +``` + +响应示例： +```json +{ + "success": true, + "content": "{\"key\": \"value\"}", + "path": "projects/config.json", + "size": 16, + "modified": 1700000000.0, + "encoding": "utf-8", + "mime_type": "application/json" +} +``` + +> 限制：最大 10MB。自动尝试 utf-8 / gbk / gb2312 / latin-1 编码。 + +--- + +## 11. 保存文件内容 + +**POST** `/api/v1/file-manager/save` + +请求方式：JSON + +| 参数 | 类型 | 位置 | 必填 | 说明 | +|------|------|------|------|------| +| path | string | body | 是 | 文件相对路径 | +| content | string | body | 是 | 文件内容 | + +```bash +curl -X POST "/api/v1/file-manager/save" \ + -H "Content-Type: application/json" \ + -d '{"path": "projects/config.json", "content": "{\"key\": \"new-value\"}"}' +``` + +响应示例： +```json +{ + "success": true, + "message": "文件保存成功", + "path": "projects/config.json", + "size": 20, + "modified": 1700000000.0, + "encoding": "utf-8" +} +``` + +> 限制：最大 5MB。保存前自动创建备份，写入失败时自动回滚。 + +--- + +## 12. 获取文件/文件夹详细信息 + +**GET** `/api/v1/file-manager/info/{file_path}` + +| 参数 | 类型 | 位置 | 必填 | 说明 | +|------|------|------|------|------| +| file_path | string | path | 是 | 文件或文件夹路径 | + +```bash +curl -X GET "/api/v1/file-manager/info/projects/report.pdf" +``` + +响应示例： +```json +{ + "success": true, + "info": { + "name": "report.pdf", + "path": "projects/report.pdf", + "type": "file", + "size": 102400, + "modified": 1700000000.0, + "created": 1699000000.0, + "permissions": "644", + "mime_type": "application/pdf", + "preview": "文件前1000字符预览..." + } +} +``` + +> 小于 1MB 的文本文件会包含 `preview` 字段。 + +--- + +## 13. 下载文件夹为 ZIP + +**POST** `/api/v1/file-manager/download-folder-zip` + +请求方式：JSON + +| 参数 | 类型 | 位置 | 必填 | 说明 | +|------|------|------|------|------| +| path | string | body | 是 | 文件夹路径 | + +```bash +curl -X POST "/api/v1/file-manager/download-folder-zip" \ + -H "Content-Type: application/json" \ + -d '{"path": "projects/dataset"}' -o dataset.zip +``` + +响应：ZIP 文件二进制流（`application/zip`）。 + +> 限制：最大 500MB，最多 10000 个文件。 + +--- + +## 14. 批量下载为 ZIP + +**POST** `/api/v1/file-manager/download-multiple-zip` + +请求方式：JSON + +| 参数 | 类型 | 位置 | 必填 | 说明 | +|------|------|------|------|------| +| paths | string[] | body | 是 | 文件/文件夹路径数组 | +| filename | string | body | 否 | ZIP 文件名，默认 `batch_download.zip` | + +```bash +curl -X POST "/api/v1/file-manager/download-multiple-zip" \ + -H "Content-Type: application/json" \ + -d '{"paths": ["projects/a.pdf", "projects/sub"], "filename": "export.zip"}' \ + -o export.zip +``` + +响应：ZIP 文件二进制流（`application/zip`）。 + +> 限制：最大 500MB，最多 10000 个文件。 + +--- + +## 15. 批量操作 + +**POST** `/api/v1/file-manager/batch-operation` + +请求方式：JSON（操作数组） + +支持操作类型：`delete` / `move` / `copy` + +```bash +curl -X POST "/api/v1/file-manager/batch-operation" \ + -H "Content-Type: application/json" \ + -d '[ + {"type": "delete", "path": "projects/old.pdf"}, + {"type": "move", "source_path": "projects/a.pdf", "target_path": "projects/archive/a.pdf"}, + {"type": "copy", "source_path": "projects/b.pdf", "target_path": "projects/backup/b.pdf"} + ]' +``` + +响应示例： +```json +{ + "success": true, + "results": [ + {"type": "delete", "success": true, "message": "删除成功"}, + {"type": "move", "success": true, "message": "移动成功"}, + {"type": "copy", "success": true, "message": "复制成功"} + ], + "total": 3, + "successful": 3 +} +``` + +--- + +## 通用说明 + +- **路径约束**：所有路径必须以 `projects` 或 `prompt` 开头，否则返回 400 +- **安全机制**：路径经过规范化处理，防止目录遍历攻击 +- **隐藏文件**：以 `.` 开头的文件/目录在列出和搜索时自动跳过 +- **错误响应格式**： +```json +{ + "detail": "错误描述信息" +} +``` From 044749e2ce5a76122ea40a677ede6dc37582a2c8 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=E6=9C=B1=E6=BD=AE?= Date: Sat, 4 Apr 2026 23:24:07 +0800 Subject: [PATCH 7/7] =?UTF-8?q?=E5=9C=A8=20"Skill=20Execution=20Workflow"?= =?UTF-8?q?=20=E4=B9=8B=E5=89=8D=E6=96=B0=E5=A2=9E=E4=BA=86=20"Skills=20vs?= =?UTF-8?q?=20Tools=20-=20CRITICAL=20DISTINCTION"=20=20=20=E6=AE=B5?= =?UTF-8?q?=E8=90=BD?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- prompt/system_prompt.md | 18 +++++++++++++++--- prompt/system_prompt_deep_agent.md | 12 ++++++++++++ 2 files changed, 27 insertions(+), 3 deletions(-) diff --git a/prompt/system_prompt.md b/prompt/system_prompt.md index 32439de..d740313 100644 --- a/prompt/system_prompt.md +++ b/prompt/system_prompt.md @@ -14,7 +14,19 @@ The filesystem backend is currently operating in: `{agent_dir_path}` - Never use relative paths in bash commands - always construct full absolute paths - Use the working directory from to construct absolute paths -**2. Skill Script Path Conversion** +**2. Skills vs Tools - CRITICAL DISTINCTION** + +**Skills are NOT tools.** Do NOT attempt to call a skill as a tool_call/function_call. + +- **Tools** (e.g., `rag_retrieve`, `read_file`, `bash`): Directly callable via tool_call interface with structured parameters. +- **Skills** (e.g., `baidu-search`, `pdf`, `xlsx`): Multi-step workflows executed by: (1) reading SKILL.md, (2) extracting the command, (3) running it via the `bash` tool. + +❌ WRONG: Generating a tool_call with `{"name": "baidu-search", "arguments": {...}}` +✅ CORRECT: Using `read_file` to read SKILL.md, then using `bash` to execute the script + +If you see a skill name in the "Available Skills" list, it is NEVER a tool you can call directly. + +**3. Skill Script Path Conversion** When executing scripts from SKILL.md files, you MUST convert relative paths to absolute paths: @@ -29,14 +41,14 @@ When executing scripts from SKILL.md files, you MUST convert relative paths to a └── scriptB.py # Actual script B file ``` -**3. Workspace Directory Structure** +**4. Workspace Directory Structure** - **`{agent_dir_path}/skills/`** - Skill packages with embedded scripts - **`{agent_dir_path}/dataset/`** - Store file datasets and document data - **`{agent_dir_path}/executable_code/`** - Place generated executable scripts here (not skill scripts) - **`{agent_dir_path}/download/`** - Store downloaded files and content -**4. Executable Code Organization** +**5. Executable Code Organization** When creating scripts in `executable_code/`, follow these organization rules: diff --git a/prompt/system_prompt_deep_agent.md b/prompt/system_prompt_deep_agent.md index 336a265..ecca407 100644 --- a/prompt/system_prompt_deep_agent.md +++ b/prompt/system_prompt_deep_agent.md @@ -83,6 +83,18 @@ When using the write_todos tool: The todo list is a planning tool - use it judiciously to avoid overwhelming the user with excessive task tracking. +### Skills vs Tools - CRITICAL DISTINCTION + +**Skills are NOT tools.** Do NOT attempt to call a skill as a tool_call/function_call. + +- **Tools** (e.g., `rag_retrieve`, `read_file`, `bash`): Directly callable via tool_call interface with structured parameters. +- **Skills** (e.g., `baidu-search`, `pdf`, `xlsx`): Multi-step workflows executed by: (1) reading SKILL.md, (2) extracting the command, (3) running it via the `bash` tool. + +❌ WRONG: Generating a tool_call with `{"name": "baidu-search", "arguments": {...}}` +✅ CORRECT: Using `read_file` to read SKILL.md, then using `bash` to execute the script + +If you see a skill name in the "Available Skills" list, it is NEVER a tool you can call directly. + ### Skill Execution Workflow **CRITICAL**: When you need to use a skill, follow this exact workflow: