Conversation
📝 WalkthroughWalkthroughDocumentation files updated to modify tool mappings guidance. "Task tool with subagents" mapping changed from unsupported stance to using Codex collab with spawn_agent and wait capability, with sequential processing as fallback when collab is disabled. Similar guidance added for subagent mentions. Changes
Estimated code review effort🎯 1 (Trivial) | ⏱️ ~3 minutes Poem
🚥 Pre-merge checks | ✅ 3✅ Passed checks (3 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing touches
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
|
Thank you1 |
…ream) - Add red flag: never start on main/master without explicit user consent - Make using-git-worktrees REQUIRED before subagent-driven-development - Make using-git-worktrees REQUIRED before executing-plans - Update integration sections with cross-references Cherry-picked from upstream PRs obra#382, obra#361
Summary
Minimal Codex-only improvement to the tool-mapping docs so Codex users understand how subagent references should be handled when collab is enabled or disabled.
Changes
.codex/superpowers-bootstrap.mdto mapTask/subagent tool references to Codex collab (spawn_agent+wait), with a sequential fallback if collab is disabled.docs/README.codex.mdto match the same guidance.Why
Codex has an experimental "Multi-agents"/"collab" feature. When enabled, Codex can use
spawn_agent. When disabled (the default), it should fall back to sequential execution. This PR makes the mapping explicit and keeps behaviorconsistent across bootstrap and documentation.
Impact
Testing
Tested in Codex with collab enabled and disabled. Behavior improved (more consistent
spawn_agentusage), but reliable multi-agent behavior still depends on explicit guidance indispatching-parallel-agentsandsubagent- driven-development, which currently requires personal overrides to be fully effective.