MCP.so
Sign In
Servers

PipeCD Docs MCP Server

@pipe-cd

A local MCP server for searching official PipeCD docs.

Overview

What is PipeCD Docs MCP Server?

PipeCD Docs MCP Server is a local MCP server that clones the official PipeCD documentation from GitHub and provides simple full-text search and document retrieval APIs via the MCP protocol. It indexes Markdown files by extracting their titles and content for use in MCP-compatible clients.

How to use PipeCD Docs MCP Server?

Install using npx after configuring your .npmrc with a GitHub PAT (with read:packages scope) and the registry @pipe-cd:registry=https://npm.pkg.github.com. Add the server configuration to your MCP client (e.g., Cursor’s mcp.json) with the command npx @pipe-cd/docs-mcp-server@latest. A Cursor deeplink is also available for one-click installation.

Key features of PipeCD Docs MCP Server

  • Full-text search across PipeCD documentation
  • Document retrieval by relative path
  • Uses sparse-checkout to minimize clone size
  • Extracts titles from Markdown front matter
  • Provides search_docs and read_docs tools
  • Simple, lightweight setup via npx

Use cases of PipeCD Docs MCP Server

  • Searching PipeCD documentation for keywords (AND search)
  • Reading the full content of a specific documentation page
  • Integrating PipeCD docs into an MCP-compatible client (e.g., Cursor)
  • Automating documentation queries in development workflows

FAQ from PipeCD Docs MCP Server

What tools does PipeCD Docs MCP Server provide?

It provides two tools: search_docs for full-text search (with parameters query, offset, limit) and read_docs for retrieving a page’s content by its relative path.

How is the documentation accessed and where is it stored?

The server clones the official PipeCD docs from GitHub into a temporary directory on first use. Markdown files are indexed there.

What are the runtime requirements?

You need Node.js/npm, and a GitHub personal access token with read:packages scope (configured in .npmrc). The server runs via stdio.

Is the search logic advanced?

No, the search logic of search_docs is simple. It does not implement advanced ranking or fuzzy matching.

What transport does the server use?

The server uses the stdio transport, invoked via the npx command. There is no built-in authentication layer; access is controlled by the MCP client configuration.

More from Memory & Knowledge