Jeo

A complete orchestration layer that runs the same four-phase workflow (plan, execute, verify, cleanup) across Claude, Codex, Gemini, and OpenCode. Uses ralph and plannotator for planning, team mode or BMAD for execution, and tracks state in a JSON file so you can restart mid-workflow. The annotate keyword triggers an optional UI feedback loop via agentation. It's ambitious plumbing for teams running multi-agent tasks across different platforms who need checkpoint recovery and don't want to rewrite orchestration logic for each one. The security audits show mixed results, so review the repo before using this in production. Good fit if you're already coordinating multiple AI agents and tired of managing the glue code yourself.

Install to Claude Code

npx -y skills add supercent-io/skills-template --skill jeo --agent claude-code