mit einem Klick
deploy-router
// Deploy claude-code-router for third-party API support (OpenRouter, Gemini, DeepSeek, etc). Use when setting up non-Anthropic API providers for TeamMCP agents.
// Deploy claude-code-router for third-party API support (OpenRouter, Gemini, DeepSeek, etc). Use when setting up non-Anthropic API providers for TeamMCP agents.
Send a file or image to WeChat via TeamMCP. Use when the user wants to send a file, image, or document to WeChat.
Send an image or screenshot to WeChat. Use when the user wants to send a picture, screenshot, or photo to WeChat.
Quick start guide for TeamMCP first-time users. Walk through installation, configuration, and first deployment in minutes.
Check and process unread TeamMCP inbox messages. Reviews all unread items, acknowledges processed ones, and summarizes what needs attention.
Generate and post a daily standup report. Checks project state, pending tasks, and unread messages, then posts a structured summary to the specified channel.
Search team knowledge across messages, project state, and task history. Compiles relevant context from multiple TeamMCP sources to answer questions.
| name | deploy-router |
| description | Deploy claude-code-router for third-party API support (OpenRouter, Gemini, DeepSeek, etc). Use when setting up non-Anthropic API providers for TeamMCP agents. |
| disable-model-invocation | true |
| allowed-tools | Bash, Read, Write, Glob, Grep |
Deploys claude-code-router as a proxy service that translates Anthropic API format to third-party providers. Required for non-Anthropic models (OpenRouter, Gemini, DeepSeek, Ollama, etc).
npx @musistudio/claude-code-router start
This installs and starts the router. Default port: 3456.
Create/edit ~/.claude-code-router/config.json:
{
"LOG": true,
"LOG_LEVEL": "info",
"HOST": "127.0.0.1",
"PORT": 3456,
"Providers": [
{
"name": "openrouter",
"api_base_url": "https://openrouter.ai/api/v1/chat/completions",
"api_key": "{YOUR_OPENROUTER_API_KEY}",
"models": ["{MODEL_NAME}"],
"transformer": {
"use": ["openrouter"]
}
}
],
"Router": {
"default": "openrouter,{MODEL_NAME}"
}
}
Provider examples:
| Provider | api_base_url | transformer |
|---|---|---|
| OpenRouter | https://openrouter.ai/api/v1/chat/completions | ["openrouter"] |
| DeepSeek | https://api.deepseek.com/chat/completions | ["deepseek"] |
| Gemini | https://generativelanguage.googleapis.com/v1beta/models/ | ["gemini"] |
| Ollama | http://localhost:11434/v1/chat/completions | (none needed) |
npx @musistudio/claude-code-router stop
npx @musistudio/claude-code-router start
curl -s -X POST http://localhost:3456/v1/messages \
-H "Content-Type: application/json" \
-H "x-api-key: test" \
-H "anthropic-version: 2023-06-01" \
-d '{"model":"{MODEL_NAME}","max_tokens":50,"messages":[{"role":"user","content":"Hello"}]}'
Should return a valid response with the model's reply.
In TeamMCP Dashboard Agent Management, set for the target agent:
auth_mode: api_keyapi_provider: provider name (e.g., openrouter)api_base_url: http://localhost:3456 (router address)api_auth_token: any value (router handles real auth)api_model: model name (e.g., qwen/qwen3-235b-a22b)Or via API:
curl -X PATCH http://localhost:3100/api/agents/{AGENT_NAME} \
-H "Authorization: Bearer {TOKEN}" \
-H "Content-Type: application/json" \
-d '{"auth_mode":"api_key","api_provider":"openrouter","api_base_url":"http://localhost:3456","api_auth_token":"test","api_model":"{MODEL_NAME}"}'
Then stop+start the agent to apply.
startAgent automatically generates these for api_key mode:
set "ANTHROPIC_API_KEY="
set "CLAUDE_CODE_OAUTH_TOKEN=channel-gate-bypass"
set "ANTHROPIC_BASE_URL=http://localhost:3456"
set "ANTHROPIC_AUTH_TOKEN=test"
set "ANTHROPIC_MODEL={MODEL_NAME}"
"transformer": {"use": ["openrouter"]} for OpenRouter compatibility