creator-analysis-mcp
Remote MCP server that provides Creator Recommendation API tools (creator_research, creator_tagged_media, etc.) with secure authentication, rate limiting, and OAuth support for Claude clients.
README
creator-analysis-mcp(Remote MCP Server)
將 Creator Recommendation API 以 remote MCP(Streamable HTTP)形式提供給 Claude 用戶。API key 只存在伺服器端,用戶只需一條 URL + bearer token。
部署到 Railway
- 將本 repo push 到 GitHub(或用
railway up直接部署) - Railway → New Project → Deploy from GitHub repo(Nixpacks 自動偵測 Node)
- 在 Variables 設定:
CREATOR_API_BASE— 後端 API base URLCREATOR_API_KEY— 後端 API key(絕不會傳給客戶端)MCP_ACCESS_TOKENS— 客戶端 token,逗號分隔,每位用戶一條(留空 = 無驗證,不建議)
- Settings → Networking → Generate Domain(或綁自訂域名,如
mcp.goodmalling.io) - Healthcheck:
/healthz
Endpoint:https://<your-domain>/mcp
用戶連接方式
Claude Code
claude mcp add --transport http creator-analysis https://<your-domain>/mcp \
--header "Authorization: Bearer <token>"
Claude Desktop / claude.ai / Cowork(自訂連接器,OAuth)
Settings → Connectors → Add custom connector → 貼上 https://<your-domain>/mcp → 按 Connect。
瀏覽器會彈出授權頁,輸入你獲發的 access token(MCP_ACCESS_TOKENS 其中一條)即完成。
OAuth session 有效 30 日;從 MCP_ACCESS_TOKENS 移除該 token 並 redeploy 即同時撤銷其所有 OAuth session。
提供的工具
creator_research、creator_tagged_media、creator_contact、creator_suggest、similar_brands
技術說明
- Stateless Streamable HTTP:每個請求獨立建立 server+transport,可水平擴展,無需 session store/Redis
- 雙重驗證:Bearer token(
MCP_ACCESS_TOKENS,支援 header 的客戶端)+ OAuth 2.0 authorization code + PKCE(claude.ai/Desktop 連接器);OAuth token 以 HMAC 簽名(OAUTH_SIGNING_SECRET)並綁定原 token 的 hash,撤銷原 token 即連帶失效 - Fail-closed:
MCP_ACCESS_TOKENS為空時拒絕所有請求(除非明確設ALLOW_OPEN=1) - Per-token rate limit:預設 30 req/min(
RATE_LIMIT_PER_MIN),超過回 429 - 上游錯誤會以
isErrortool result 回傳,不會洩漏 key - 本地開發:
cp .env.example .env填好後npm install && npm start,endpoint 在http://localhost:8080/mcp
建議後續
- 在 Railway 對此服務設 usage alert;上游 embedding 服務建議加 keep-warm 避免 502 冷啟動
- 若要按用戶計量/計費,可在
authorized()內按 token 記錄用量
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.