---

name: brainerd description: "Use Brainerd as the umbrella router for repo-memory actions across Pi, Codex, and Claude. Repos that already have a brain may load Brainerd ambiently without invoking this skill." license: MIT

Brainerd

Brainerd may already be ambient in repos that have a brain. In that case the active harness should already be reading brain/index.md and brain/principles.md before non-trivial work.

Use this root skill only to route to the correct harness-specific surface.

Canonical surfaces

• Pi: pi-init, pi-reflect, pi-ruminate
• Codex: codex-init, codex-reflect, codex-ruminate
• Claude: claude-init, claude-reflect, claude-ruminate

Routing rules

• If the repo already has a Brainerd brain, you do not need this skill just to make the harness read it.
• If the user wants first-time setup, use the matching *-init surface for the current harness.
• If the user wants to preserve something from the current conversation, use the matching *-reflect surface.
• If the user wants to mine older repo-scoped session history, use the matching *-ruminate surface.
• Do not use the legacy names brainerd-init, brainerd-reflect, brainerd-ruminate, /brain-init, /reflect, or /ruminate.

Shared rules

• Edit only under brain/ plus the managed Brainerd block in AGENTS.md.
• Do not hand-edit generated entrypoints like brain/index.md or brain/principles.md.
• Claude imports belong under brain/imports/claude/; distill them into notes or principles instead of copying them verbatim.
• End with a short summary of what changed, what was only previewed, or why no brain changes were written.