sentry-mcp

Tech stack & purpose

Sentry MCP is a Model Context Protocol server that exposes Sentry's error tracking and performance monitoring capabilities to AI assistants through 19 tools. Built with TypeScript, the project comprises multiple packages including a core MCP implementation, stdio transport, Cloudflare web app with OAuth support, and test infrastructure. The server employs a skills-based authorization system that maps user-friendly capabilities to technical API permissions, with skills like inspect, seer, triage, and project-management controlling access to different tool subsets. AI-powered search tools can leverage configured LLM providers such as OpenAI or OpenRouter to translate natural language queries into Sentry's query syntax, and the project emphasizes type safety, security practices, and comprehensive testing with strict quality gates requiring TypeScript compilation, linting, and test passage before commits.

Languages

Contributors (top 30 of 65)

• David Cramer892 commits
• getsentry-bot75 commits
• sentry-junior[bot]42 commits
• Cody De Arkland11 commits
• Cameron Cooke8 commits
• sentry-release-bot[bot]7 commits
• Sergiy Dybskiy7 commits
• PickleNik6 commits
• geoffg-sentry6 commits
• Michi Hoffmann6 commits
• Elio Di Nino5 commits
• Nico Hinderling5 commits
• Jan Peer Stöcklmair5 commits
• seer-by-sentry[bot]5 commits
• 
  dependabot[bot]5 commits · 0 followers
• William Mak5 commits
• Michael Hoffmann5 commits
• Miguel Betegón5 commits
• 石岳峰4 commits
• Abdullah Khan3 commits
• Daniel Griesser3 commits
• Burak Yigit Kaya3 commits
• sentry[bot]3 commits
• Henry Mao3 commits
• John Manhart3 commits
• Shaun Kaasten3 commits
• Vjeran Grozdanić3 commits
• ɥsoɾ2 commits
• Abhijeet Prasad2 commits
• Matt Duncan2 commits