Automated GitHub issue resolution that spins up parallel analysis agents to investigate root cause, search for similar patterns in your codebase and memory MCP, then generates a fix with tests and prevention recommendations. It fetches the issue via gh CLI, scales investigation depth by effort level (quick fix vs full RCA), and can resume mid-flow if interrupted. The task management is mandatory and well-structured, though the docs lean heavy on process. Best for teams that want systematic debugging with traceable reasoning instead of one-off patches. Requires memory and context7 MCP servers plus some setup, but the multi-host support for GitHub Enterprise, GitLab, and Bitbucket is solid if you work across platforms.
npx skills add https://github.com/yonatangross/orchestkit --skill fix-issue