home / mcp / ecosyste-ms mcp server
Queries a local critical-packages.db and provides MCP endpoints for ecosystem data and tooling.
Configuration
View docs{
"mcpServers": {
"ecosyste-ms-mcp": {
"command": "npx",
"args": [
"@ecosyste-ms/mcp"
],
"env": {
"ECOSYSTEMS_DB_PATH": "YOUR_PATH_TO_DB (optional)"
}
}
}
}You deploy this MCP server to quickly query ecosystem data from ecosyste.ms. It reads a local SQLite database for fast lookups and can fall back to an API if a package isn’t in the database, making it a fast and flexible data source for your tooling and LLM workflows.
You run the MCP server locally and connect your MCP client to it. Use the standard MCP command to start the server, then configure your client to point at the local or tunnel URL. Once running, you can query package data, search packages, retrieve versions and advisories, and inspect repository metadata. You can also verify database health and check which registries are available.
Prerequisites: Node.js and npm should be available on your system.
npm install -g @ecosyste-ms/mcpOr run directly with npx to avoid global installation.
npx @ecosyste-ms/mcpThe database is bundled via @ecosyste-ms/critical. No additional setup required.
To use a custom database, set ECOSYSTEMS_DB_PATH or place critical-packages.db in your working directory.
# Example of setting a custom database path when starting via a client
ECOSYSTEMS_DB_PATH=/path/to/critical-packages.db npx @ecosyste-ms/mcpYou can run the MCP server through different clients or editors. The following example shows a standard stdio-based configuration you would load in your MCP client setup.
{
"mcpServers": {
"ecosystems": {
"type": "stdio",
"name": "ecosystems",
"command": "npx",
"args": ["@ecosyste-ms/mcp"]
}
}
}If you need to point the server at a specific database, provide the path via an environment variable when starting the server through your MCP client.
ECOSYSTEMS_DB_PATH=/path/to/critical-packages.db npx @ecosyste-ms/mcpThe server is designed to be used with MCP clients and integrates with LLM tools and editors to make data access seamless. You’ll typically run the server via a simple command and keep it running while your tooling queries the local database or falls back to the API when needed.
Get full package data by ecosystem and name
Find a package by PURL, ecosystem+name, or repository URL
List all versions with publish dates for a package
Retrieve security advisories (CVEs) for a package
Fetch repository metadata such as stars, forks, and language
Identify packages that depend on a given package
Perform a full-text search against the local database (requires the local DB)