The best Pi coding agent setup — 10 packages, 59 skills, 10-step autonomous workflow. Zero bloat, zero MCP, pure Pi ideology.
Skill Claude CodeCodex
Generate a "pick up where I left off" status report across all active Pi sessions. Use when the user asks for bearings, a status report, morning brief, catch-up, "where did I leave off", or "what's in the works". Reads live session state, composes a scannable 4-section digest, and writes a dated report to…
Skill Claude CodeCodex
You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.
Skill Claude CodeCodex
Review the changes since a fixed point (commit, branch, tag, or merge-base) along two axes — Standards (does the code follow this repo's documented coding standards?) and Spec (does the code match what the originating issue/spec asked for?). Runs both reviews in parallel sub-agents and reports them side by side. Use…
Skill Claude CodeCodex
Find semantic duplicates (same intent, different implementation) and shallow modules (thin wrappers spreading complexity). Read-only advisory — changes route through BUILD with full gates. Use when auditing code quality, before refactoring, or when a codebase feels bloated.
Skill Claude CodeCodex
Never lose an unresolved decision. Use when work discovers an open question that needs the user's input, before declaring that work complete. Persists decisions to /.pi/agent/decisions.json so they survive compaction and session restarts. Inspired by FirstMate's decision-hold-lifecycle.
Skill Claude CodeCodex
Diagnosis loop for hard bugs and performance regressions. Use when the user says "diagnose"/"debug this", or reports something broken/throwing/failing/slow.
Skill Claude CodeCodex
Treat documentation as a first-class deliverable of every BUILD phase. Classify the diff's doc impact across 3 layers (business/technical/audit) and write only the updates genuinely needed. Use after BUILD completes, before a commit/PR, or when asked "are docs up to date".
Skill Claude CodeCodex
Review and sharpen persistent memory so it compounds instead of accumulating. Use when pruning pi-hermes-memory entries, doing monthly memory hygiene, or when the same lesson has been recorded multiple times.
Skill Claude CodeCodex
The auto-pi workflow procedure layer. ask-matt owns semantic routing; the deterministic workflow interpreter owns state transitions, evidence gates, persistence, autonomous continuation, bounded retries, review, and ship.
Skill Claude CodeCodex
Use when receiving code review feedback, before implementing suggestions, especially if feedback seems unclear or technically questionable - requires technical rigor and verification, not performative agreement or blind implementation.
Skill Claude CodeCodex
Security review of the diff — the third review axis alongside Standards and Spec. Reviews for injection, auth, secrets, deserialization, SSRF, path traversal, unsafe operations, and dependency confusion. Use as a parallel sub-agent with fresh context (diff-only, anti-anchored) during the review phase.
Skill Claude CodeCodex
Hand off work to another Pi session — write a handoff doc, save key decisions to memory, and notify the target session via intercom. Use when a session is getting long, when you need to continue work in a fresh context, or when transferring work between projects/sessions. Composes handoff.ts (doc writer) + memory…
Skill Claude CodeCodex
Maintain and improve a Pi coding agent setup over time — prevent drift AND find new leverage. Use when auditing setup health, before adding a package/extension, before publishing, or when asked "is my setup healthy" / "audit my setup" / "how do I keep this sharp".
Skill Claude CodeCodex
Use when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence before assertions always.