rig

Use when a user wants to use Rig to make software development more efficient through delegated analysis, focused reviews, isolated patch runs, artifact inspection, subagent coordination, parallel task decomposition, or deciding when to delegate work versus editing directly.

Use when implementing or reviewing Rig itself, including changes to the CLI, runners, run artifacts, worktree isolation, MCP adapter, environment diagnostics, or agent harness documentation.

Monitor and harden GitHub pull requests until they are ready to merge. Use when Codex is asked to watch a PR, fix failing CI, inspect GitHub Actions failures, respond to CodeRabbit or other review bot comments, address human review feedback, re-run checks, resolve merge blockers, or bring a PR to merge-ready state.