What tools does Scout Mcp provide?
It offers 21 tools including `hackernews_search`, `github_search`, `npm_search`, `pypi_search`, `x_search`, `reddit_search`, `youtube_search`, `arxiv_search`, and `scout_report` for parallel multi-source searches.
@bartonguestier1725-collab
基本情報
概要はまだありません
以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。
{
"mcpServers": {
"scout": {
"command": "node",
"args": [
"/path/to/scout-mcp/build/index.js"
]
}
}
}Hacker News (Algolia)
GitHub
GitHub
npm Registry
PyPI
Product Hunt
X / Twitter
Dev.to
Hashnode
Lobste.rs
StackExchange
ArXiv
YouTube
Zenn
Qiita
Semantic Scholar
Lemmy
GitLab
x402 Bazaar
Multi-source
Scout Mcp is an MCP server that searches 21 platforms in parallel and returns structured JSON. It integrates with code registries, academic papers, social platforms, and tech blogs, and can be connected to any MCP client such as Claude Desktop, VS Code, or Cursor.
Clone the repository, install dependencies with npm install, and build with npm run build. Add the server configuration to your MCP client’s config file with the command node and the path to the built index.js. Optionally run via Docker using docker build -t scout-mcp . and docker run -i scout-mcp.
Promise.allSettled.It offers 21 tools including `hackernews_search`, `github_search`, `npm_search`, `pypi_search`, `x_search`, `reddit_search`, `youtube_search`, `arxiv_search`, and `scout_report` for parallel multi-source searches.
No, most tools work without any keys. Optional keys are required only for `x_search`, `youtube_search`, `producthunt_search`, and `reddit_search`, while others like GitHub, Qiita, Semantic Scholar, and StackExchange have optional keys to raise rate limits.
Scout Mcp uses the standard stdio transport, reading JSON-RPC messages from stdin and writing responses to stdout.
Each tool makes a direct API call to its respective source (e.g., GitHub API, npm registry, ArXiv API) and returns structured JSON with fields like `title`, `url`, `source`, and `created_at`.
The server is built with the Model Context Protocol SDK and requires Node.js. The Dockerfile uses `node:24-alpine`. All configuration is done via environment variables.

Handl speaks the Model Context Protocol. Your whole billing operation — projects, invoices, clients, reports, approvals — as tools any AI assistant can call, inside your autonomy rules.

DeepMark helps teachers deliver rapid, consistent marking with meaningful feedback for every student — in a fraction of the time. What once took a week, now takes one free period.

See and control a real Windows PC you own — from any MCP client, locally or remotely. Observe (UIA + screenshots), click/type/drag/scroll, launch apps, owner Live View. BYOH: your machine, your key.
MCP Toolbox for Databases is an open source MCP server for databases.

Connect your AI assistant to your own YOUB endurance training account. Read recorded activities with full analysis, planned sessions, race goals, wellness (HRV, sleep, resting HR) and your athletic profile. Create or mov
コメント