--- title: "MCP 工具集" url: "/docs/mcp_toolsets" canonical_url: "https://docs.litellm.ai/docs/mcp_toolsets" type: "docs" last_updated: "2026-10-08" related: - "/docs/mcp_server_submissions" - "/docs/proxy/tool_policies" --- # MCP 工具集 > Index of all LiteLLM docs: https://docs.litellm.ai/llms.txt 如需直接用戶端 URL 與 LLM 工具選擇器，請參閱 [MCP 組態參考](./mcp_config_reference#endpoint-matrix) **Toolset** 是由一或多個 MCP 伺服器中的特定工具所組成的具名集合。您不必讓代理程式存取每個伺服器上的所有工具，而是可以精確選取它實際需要的工具，無論這些工具位於哪個伺服器，並將它們捆綁在單一名稱下。 ## 運作方式 {#how-it-works} ``` ┌─────────────────────────────────┐ │ MCP Toolset │ │ "devtooling-prod" │ └────────────┬────────────────────┘ │ ┌──────────────────┴──────────────────┐ │ │ ┌────────▼────────┐ ┌────────▼────────┐ │ CircleCI MCP │ │ DeepWiki MCP │ │ (10+ tools) │ │ (3 tools) │ └────────┬────────┘ └────────┬────────┘ │ │ ┌─────────┴──────────┐ ┌──────────┴──────────┐ │ ✓ get_build_logs │ │ ✓ read_wiki_structure│ │ ✓ find_flaky_tests │ │ ✓ read_wiki_contents │ │ ✓ get_pipeline_ │ │ ✗ ask_question │ │ status │ └─────────────────────┘ │ ✓ run_pipeline │ │ ✗ list_followed_ │ │ projects │ └────────────────────┘ Agent sees exactly 6 tools, nothing more. ``` 原本跨兩個伺服器有 13+ 個工具，代理程式只會取得 6 個：也就是它實際需要的那些。 **這為何重要：** - 較短的工具清單 → 更少 token、更快的回應、更少幻覺 - 將 GitHub + Linear + CircleCI 的工具組合成一個具名授權 - 以相同方式指派給金鑰和團隊，就像您今天指派 MCP 伺服器一樣 --- ## 建立工具集 {#create-a-toolset} ### 1. 前往 MCP 頁面 {#1-go-to-the-mcp-page} 在左側側邊欄中導覽至 **MCP**。 ![導覽至 MCP](https://colony-recorder.s3.amazonaws.com/files/2026-03-22/1a96c713-6a37-4f96-92f1-07bd58c1973c/ascreenshot_23515f386ccc4597b0633987667fe01f_text_export.jpeg) ### 2. 開啟 Toolsets 分頁 {#2-open-the-toolsets-tab} 在 MCP 頁面上點擊 **Toolsets** 分頁。 ![點擊 Toolsets 分頁](https://colony-recorder.s3.amazonaws.com/files/2026-03-22/65b6986b-595a-4b28-8fdc-a7b36bc76e59/ascreenshot_ca70c18fe7ec415486f96a6b405bf550_text_export.jpeg) ### 3. 點擊 "New Toolset" {#3-click-new-toolset} ![New Toolset 按鈕](https://colony-recorder.s3.amazonaws.com/files/2026-03-22/798c55c4-5d6b-4815-a642-70ac9f34f102/ascreenshot_3f144f54a1a944e28454239c837b4e6d_text_export.jpeg) ### 4. 輸入名稱 {#4-enter-a-name} 為 toolset 輸入名稱。請選擇具有描述性的名稱，因為代理程式會引用這個名稱。 ![輸入工具集名稱](https://colony-recorder.s3.amazonaws.com/files/2026-03-22/62b412e0-d38f-44c3-99e4-3693f1512f6a/ascreenshot_b678c7c988a04f8b887b0f54c4dd95a7_text_export.jpeg) ![工具集名稱欄位](https://colony-recorder.s3.amazonaws.com/files/2026-03-22/ba5ebc95-cab7-470b-a7c9-21f12b9b01a3/ascreenshot_a602e982a2a44890a83dca64d61c38eb_text_export.jpeg) ### 5. 新增第一個工具 {#5-add-the-first-tool} 從下拉選單中選取一個 MCP 伺服器，然後選擇要從該伺服器納入的工具。 ![選取 MCP 伺服器](https://colony-recorder.s3.amazonaws.com/files/2026-03-22/2aa5bcba-6414-42e3-9813-efb0a9078e32/ascreenshot_58fbff35ba654210a1b4dc5452aa6bd9_text_export.jpeg) ![從下拉選單中選擇伺服器](https://colony-recorder.s3.amazonaws.com/files/2026-03-22/4fd9cffb-d3ba-461a-8679-89f278bf67ad/ascreenshot_b61e9e85a51b494a8d09fe61198d63e1_text_export.jpeg) ![從伺服器選取工具](https://colony-recorder.s3.amazonaws.com/files/2026-03-22/60718e72-2062-494b-9a23-456992c88cbd/ascreenshot_7a1f8eeab30a4a05ba39c450e5458b78_text_export.jpeg) ### 6. 新增來自第二個伺服器的工具 {#6-add-tools-from-a-second-server} 按一下 **Add Tool**，選擇不同的 MCP 伺服器，然後再選取另一個工具。依需要重複，工具可以來自任意數量的伺服器。 ![從第二個伺服器新增工具](https://colony-recorder.s3.amazonaws.com/files/2026-03-22/f34e0600-cc74-4b18-8794-88d45f326144/ascreenshot_98834b14ab9343e39fb503e458d72b7c_text_export.jpeg) ![選取第二個伺服器](https://colony-recorder.s3.amazonaws.com/files/2026-03-22/75150368-2202-4da1-99f1-6f0620e9b133/ascreenshot_f94d0bc08ea147348a9cf021cce7d854_text_export.jpeg) ![從第二個伺服器選取工具](https://colony-recorder.s3.amazonaws.com/files/2026-03-22/ed2cdf6e-025d-4d50-8b12-ed68745d5c51/ascreenshot_0c1c7f76524b46c5a056fda5e6956e2b_text_export.jpeg) ### 7. 建立工具集 {#7-create-the-toolset} 點擊 **Create Toolset** 以儲存。 ![建立工具集](https://colony-recorder.s3.amazonaws.com/files/2026-03-22/021ca7b3-2d9a-49a0-8758-dae3dc3bcb4d/ascreenshot_14c6434e71114a6091e359a996f20e12_text_export.jpeg) --- ## 在 Playground 中使用工具集 {#use-a-toolset-in-the-playground} 建立後，您的 toolset 會與 MCP 伺服器一起出現在 Playground 的 **MCP Servers** 下拉選單中，並可用相同方式選取。 ### 1. 前往 Playground {#1-go-to-the-playground} ![導覽至 Playground](https://colony-recorder.s3.amazonaws.com/files/2026-03-22/f9d4aa4c-d98e-4767-b98e-aad2890e97ca/ascreenshot_d84239c441bb4e828f229d0c9e079e3f_text_export.jpeg) ![點擊 Playground](https://colony-recorder.s3.amazonaws.com/files/2026-03-22/d8a07563-97fe-453a-b974-88da46c87294/ascreenshot_ea494300a536400abb2ea6bf3bdfd5ab_text_export.jpeg) ### 2. 從 MCP Servers 中選取您的工具集 {#2-select-your-toolset-from-mcp-servers} 在左側面板的 **MCP Servers** 下，開啟下拉選單並選取您的工具集。模型只會看到您納入其中的工具。 ![選取 MCP servers 下拉選單](https://colony-recorder.s3.amazonaws.com/files/2026-03-22/ee8cb38c-c4ff-4b4b-844c-22f2e40832ae/ascreenshot_e300fb39cea0434fb5e3986e912a2b8d_text_export.jpeg) ![開啟 MCP 伺服器選擇器](https://colony-recorder.s3.amazonaws.com/files/2026-03-22/8672070c-5d07-4f63-878c-6fc7dcbc9b65/ascreenshot_326ddd0868224c99a6fa5dab2d144f1f_text_export.jpeg) ![選取工具集](https://colony-recorder.s3.amazonaws.com/files/2026-03-22/955826ad-2bbb-403e-ab26-c1ac03ec2675/ascreenshot_13f837ad53574535986ca7ca5998d34a_text_export.jpeg) ![已選取並啟用工具集](https://colony-recorder.s3.amazonaws.com/files/2026-03-22/9a59c3b9-1563-4731-838f-1c35d636ddc9/ascreenshot_c05d8fa5f37a4b3093fc46e26f293b4d_text_export.jpeg) 模型現在只能存取您工具集中的工具，沒有其他工具。 --- ## 透過 API 使用工具集 {#use-a-toolset-via-api} 將 toolset 的 route 作為 tools 清單中的 `server_url` 傳入。LiteLLM 會在伺服器端解析，因此不需要公開 URL。 **Responses API** ```python import openai client = openai.OpenAI( api_key="your-litellm-key", base_url="http://your-proxy/v1", ) response = client.responses.create( model="gpt-5.6-terra", input="What CI/CD tools do you have?", tools=[ { "type": "mcp", "server_label": "devtooling-prod", "server_url": "litellm_proxy/mcp/devtooling-prod", "require_approval": "never", } ], ) print(response.output_text) ``` **Chat Completions API** ```python import openai client = openai.OpenAI( api_key="your-litellm-key", base_url="http://your-proxy/v1", ) response = client.chat.completions.create( model="gpt-5.6-terra", messages=[{"role": "user", "content": "What CI/CD tools do you have?"}], tools=[ { "type": "mcp", "server_label": "devtooling-prod", "server_url": "litellm_proxy/mcp/devtooling-prod", "require_approval": "never", } ], ) print(response.choices[0].message.content) ``` **REST** ```bash curl http://your-proxy/v1/responses \ -H "Authorization: Bearer your-litellm-key" \ -H "Content-Type: application/json" \ -d '{ "model": "gpt-5.6-terra", "input": "What CI/CD tools do you have?", "tools": [ { "type": "mcp", "server_label": "devtooling-prod", "server_url": "litellm_proxy/mcp/devtooling-prod", "require_approval": "never" } ] }' ``` --- ## 透過 API 管理工具集 {#manage-toolsets-via-api} ```bash # List all toolsets curl http://your-proxy/v1/mcp/toolset \ -H "Authorization: Bearer your-litellm-key" # Create a toolset curl -X POST http://your-proxy/v1/mcp/toolset \ -H "Authorization: Bearer your-litellm-key" \ -H "Content-Type: application/json" \ -d '{ "toolset_name": "devtooling-prod", "description": "CircleCI + DeepWiki tools for the dev team", "tools": [ {"server_id": "", "tool_name": "get_build_failure_logs"}, {"server_id": "", "tool_name": "run_pipeline"}, {"server_id": "", "tool_name": "read_wiki_structure"} ] }' # Delete a toolset curl -X DELETE http://your-proxy/v1/mcp/toolset/ \ -H "Authorization: Bearer your-litellm-key" ``` ## Related pages - [MCP 伺服器提交](https://docs.litellm.ai/docs/mcp_server_submissions.md) - [工具政策](https://docs.litellm.ai/docs/proxy/tool_policies.md)