cratesdotio-mcp

Crates.iocratesdotio-mcp
lib.rscratesdotio-mcp
version0.1.0
created_at2026-01-09 14:12:35.376605+00
updated_at2026-01-09 14:12:35.376605+00
descriptionA Model Context Protocol (MCP) server for searching and retrieving crate information from crates.io
homepage
repositoryhttps://github.com/fedemagnani/crates.io-mcp
max_upload_size
id2032150
size82,821
Federico Magnani (fedemagnani)

documentation

README

cratesdotio-mcp

MCP server for interacting with crates.io.

This mcp tool can be useful to define rules on your AI agent in order to:

  • Use always the latest version of a crate
  • Retrieve the github repository of a crate
  • Access brief description of what a crate does
  • Access dependencies of a crate

Tools

  • search_crates - Search crates by name/keyword
  • get_crate_info - Get crate details
  • get_crate_versions - List published versions
  • get_crate_dependencies - Show dependencies for a version
  • get_crate_documentation - Get docs.rs links

Installation

cargo build --release

Running

Stdio Mode (default)

./target/release/cratesdotio-mcp

HTTP Mode

# Default port 3000
./target/release/cratesdotio-mcp --http

# Custom port
./target/release/cratesdotio-mcp --http 8080

Docker

docker build -t cratesdotio-mcp .

# HTTP mode (custom port)
docker run -d -p 8080:8080 cratesdotio-mcp --http 8080

# Stdio mode
docker run -i cratesdotio-mcp

MCP Configuration

Make sure that the mcp server is running BEFORE starting/configuring your AI agent in the editor: this will make sure that the initial session handshake is successful.

HTTP (local or Docker)

{
  "mcpServers": {
    "crates-io": {
      "serverUrl": "http://localhost:3000/sse"
    }
  }
}

Stdio (local binary)

{
  "mcpServers": {
    "crates-io": {
      "command": "/path/to/cratesdotio-mcp"
    }
  }
}
Commit count: 10

cargo fmt