DeepSource MCP Server

Name: DeepSource MCP Server
Author: sapientpants

@sapientpants

リポジトリ

DeepSource MCP Server について

Model Context Protocol (MCP) server for DeepSource

#typescript #mcp #deepsource #model-context-protocol #mcp-server #model-context-protocol-servers

基本情報

カテゴリ

その他

ライセンス

MIT license

ランタイム

node

トランスポート

stdio

公開者

sapientpants

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": {
    "deepsource": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "DEEPSOURCE_API_KEY",
        "sapientpants/deepsource-mcp-server"
      ],
      "env": {
        "DEEPSOURCE_API_KEY": "your-deepsource-api-key"
      }
    }
  }
}

ツール

No required parameters

`projectKey` (required) - The unique identifier for the DeepSource project

`runIdentifier` (required) - The runUid (UUID) or commitOid (commit hash) to identify the run

概要

What is DeepSource MCP Server?

The DeepSource MCP Server is a Model Context Protocol (MCP) server that integrates with DeepSource to give AI assistants access to code quality metrics, issues, and analysis results. It enables AI assistants to retrieve code metrics, filter issues, check quality status, and analyze project quality over time via DeepSource’s GraphQL API.

How to use DeepSource MCP Server?

Install via npm (npx [email protected]) or run the Docker image (sapientpants/deepsource-mcp-server). Provide your DeepSource API key as the environment variable DEEPSOURCE_API_KEY. Configure the server in your AI assistant’s MCP client, e.g., by adding a deepsource entry to claude_desktop_config.json with the appropriate command and environment variable.

Key features of DeepSource MCP Server

Integrates with DeepSource via GraphQL API
Implements the Model Context Protocol for AI assistants
Retrieves project lists, issues, and analysis runs
Supports pagination and filtering of results
Built with TypeScript for type safety
Runs on Linux, macOS, and Windows

Use cases of DeepSource MCP Server

An AI assistant checks the latest code quality issues from a DeepSource project
A developer queries analysis run history for a specific project
An automated workflow retrieves issue details for a given commit
A code review tool surfaces DeepSource findings during pull request analysis

FAQ from DeepSource MCP Server

What does the DeepSource MCP Server do?

It acts as a bridge between AI assistants (via the Model Context Protocol) and DeepSource’s code quality analysis platform, allowing assistants to query projects, issues, and analysis runs.

What are the prerequisites for running this server?

Node.js 20 or higher and pnpm 10.7.0 or higher for local builds, or Docker for containerized deployment. You also need a DeepSource API key.

How do I provide my DeepSource API key?

Set the environment variable DEEPSOURCE_API_KEY when starting the server. Examples are shown in the configuration snippets for Docker and npx.

What tools does the server expose?

Four tools: deepsource_projects (list projects), deepsource_project_issues (get issues with optional pagination), deepsource_project_runs (list analysis runs with pagination), and deepsource_run (get a specific run by runUid or commitOid).