MCP.so
ログイン

Bucketeer Docs Local MCP Server

@bucketeer-io

基本情報

カテゴリ

メモリとナレッジ

ライセンス

Apache-2.0

トランスポート

stdio

作者

bucketeer-io

Bucketeer Docs Local MCP Server について

Local MCP Server to query Bucketeer documentation

Bucketeer Docs Local MCP Server の設定

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": {
    "bucketeer-docs-local-mcp-server": {
      "command": "npx",
      "args": [
        "@bucketeer/docs-local-mcp-server",
        "build-index"
      ]
    }
  }
}

Bucketeer Docs Local MCP Server のツール

ツール一覧はまだありません

この掲載には表示できるツール一覧がまだありません。サーバーにツールがないという意味ではありません。対応機能はプロジェクトのドキュメントをご確認ください。

上のボタンで一時的なサンドボックスでこのサーバーを実行し、ツール一覧を取得できます。

概要

What is Bucketeer Docs Local MCP Server?

A Model Context Protocol (MCP) server that provides search and retrieval of Bucketeer’s feature flag and experimentation platform documentation. It enables AI assistants to answer questions about Bucketeer’s features and usage by indexing and serving content from the official documentation repository.

How to use Bucketeer Docs Local MCP Server?

Install dependencies (npm install), build the project (npm run build), and build the document index (npm run build:index). Start the server with npm start or run it via npx @bucketeer/docs-local-mcp-server. Configure the server in your MCP client (e.g., Cursor or Claude Desktop) by adding the npx command to its mcpServers configuration or by using the provided Cursor install deeplink.

Key features of Bucketeer Docs Local MCP Server

  • Automatically fetches .mdx files from the GitHub repository.
  • Intelligent indexing with keyword extraction and relevance scoring.
  • Caches documents locally using SHA‑based change detection.
  • Two exposed tools: search_docs and get_document.
  • Configurable default search result limit (default: 5).
  • Supports recursive directory traversal for all documentation subdirectories.

Use cases of Bucketeer Docs Local MCP Server

  • AI assistants answering user questions about Bucketeer feature flags.
  • Quickly finding and retrieving specific documentation sections on SDKs, experiments, or targeting.
  • Integrating Bucketeer documentation context into development tools like Cursor or Claude Desktop.

よくある質問

1

What are the system requirements?

Node.js 18+ and npm are required to run the server.

2

Where does the documentation data come from?

It fetches `.mdx` files from the `docs/` directory of the [bucketeer-io/bucketeer-docs](https://github.com/bucketeer-io/bucketeer-docs) repository on GitHub.

3

How are documents cached?

Files are cached locally as JSON in `files/docs/`, the document index is stored in `files/index/document-index.json`, and a GitHub cache with SHA‑based change detection is kept in `files/docs/github_cache.json`.

4

How do I update the documentation index?

Run `npm run build:index` or `npx @bucketeer/docs-local-mcp-server build-index`. Use the `--force` flag to ignore the cache and rebuild everything.

5

What tools does this server provide?

Two tools: `search_docs` (accepts a query string and optional limit) returns ranked documentation snippets, and `get_document` (accepts a document path) returns the full content of that document.

コメント

「メモリとナレッジ」の他のコンテンツ