File Search MCP

What is File Search MCP?

File Search MCP is a specialized Model Context Protocol (MCP) server that provides full‑text search capabilities for text files in a given directory. It is built with Rust and uses the Tantivy search engine for efficient indexing and querying. The server is meant for AI assistants and other MCP‑compatible systems that need to search local file contents.

How to use File Search MCP?

First, install the Rust toolchain. Clone the repository (git clone [email protected]:Kurogoma4D/file-search-mcp.git), then compile the binary. Add the server to your MCP settings (e.g., in Cursor or Claude) by specifying the command as <path‑to‑repo>/target/release/file‑search‑mcp.

Key features of File Search MCP

• Full‑text search across a directory structure
• File content reader to view specific text files
• Smart detection of text files, skipping binaries
• MCP integration for AI assistant compatibility
• In‑memory indexing for temporary, fast searches
• Score‑based results with relevance rankings

Use cases of File Search MCP

• Searching a codebase for variable names or error messages
• Reading the contents of a specific file through an AI assistant
• Integrating full‑text search into a Claude or Cursor workflow
• Quickly scanning documentation files for keywords

FAQ from File Search MCP

What files can File Search MCP search?

Only text files are indexed and searched; binary files are automatically skipped.

How does the indexing work?

The server creates an in‑memory Tantivy index from the text files in the specified directory. The index is temporary and is rebuilt on each run.

What are the parameters of the search tool?

The search tool requires two parameters: directory (path to the directory to search) and keyword (keyword to search for).

Can I read the content of a file directly?

Yes, the file content reader tool accepts a file_path parameter and displays the file’s content.

What license is File Search MCP released under?

MIT License.