Open-source MCP server that generates a data-grounded sprint retrospective from your Jira, GitHub, and Slack activity. What went well, what did not, and action items, each backed by evidence.
Open source on GitHub and published on npm.
Retros usually run on memory, so they run on bias: the loudest voice wins, the same complaints recur every sprint, and the action items quietly never happen. retro-mcp reads the actual record of the sprint instead. You ask your AI assistant for the retro and it answers with what went well, what did not, and action items, every line backed by a specific metric, ticket, PR, or Slack thread. It frames everything around the process, never individuals, and it flags the problems that keep recurring across sprints.
It is a Model Context Protocol server, so it works with any MCP client (Claude, Cursor, Cline, and more) on any model. It is local and read-only, so nothing about your sprint leaves your machine except calls to your own tools' APIs, and it needs no AI API key. A retro aid, not a surveillance tool. It also ships a zero-config demo mode.
Try it.npx -y retro-mcp --demo
Stack. TypeScript, Model Context Protocol SDK, Jira Cloud REST, GitHub REST, and Slack Web APIs, changelog-derived metrics, tested end to end. Built with Claude Code.
Your email address will not be published. Required fields are marked *