Agent Claude Code
Analyze requirements, ambiguity, and hidden constraints When NOT to use: structuring the how (planner); post-implementation verification (verifier).
Agent Claude Code
Independently re-verify findings WITHOUT trusting prior analysis — an unbiased cold check to catch confirmation bias the chained reviewer/verifier path can introduce When NOT to use: routine correlation verification (verifier — faster and cheaper); first-pass review.
Agent Claude Code
Challenge plans and designs before execution When NOT to use: reviewing implemented code (reviewer); trivial one-phase plans with no risk surface.
Agent Claude Code
Implement planned code changes When NOT to use: design/UX decisions (designer), architecture calls (oracle), exploration-heavy discovery (explorer).
Agent Claude Code
Fast codebase discovery and file/symbol mapping When NOT to use: write-actions of any kind; external docs lookup (librarian); deep adversarial analysis (council).
Agent Claude Code
Create an execution plan with clear sequencing and risk notes When NOT to use: requirements clarification (analyst); executing phases (executor).
Agent Claude Code
Review code changes for correctness, maintainability, and regressions When NOT to use: pre-execution plan critique (critic); security threat-modeling (security-reviewer).
Agent Claude Code
Review changes for security vulnerabilities and trust-boundary issues When NOT to use: general correctness/style review (reviewer); running audits yourself — no shell access, route audit commands to executor.
Agent Claude Code
Design and implement test strategy for a change When NOT to use: implementing the feature under test (executor); one-off verification runs (verifier).
Agent Claude Code
Verify that implementation satisfies the requested goal When NOT to use: adversarial cold re-checks of high-stakes verdicts (cold-verifier); authoring new tests (test-engineer).
Agent Claude Code
Write concise documentation, migration notes, and summaries When NOT to use: code changes (executor); verifying claims — mark sections DRAFT instead.
Skill Claude CodeCodex
Background worker, heartbeat, stale-run, crash-recovery, and deadletter workflow. Use when debugging stuck/dead workers or changing async run reliability. When NOT to use: pure investigation without async context (use read-only-explorer); one-shot fixes that don't need recovery infrastructure.
Skill Claude CodeCodex
Child Pi worker spawning, lifecycle callbacks, and failure modes. When NOT to use: non-Pi subprocess execution (use safe-bash); for general async reliability (use async-worker-recovery).
Skill Claude CodeCodex
Use when constructing worker prompts, reading artifacts/logs, summarizing runs, compacting context, or handing work between agents. When NOT to use: tasks without artifact/context concerns; pure code questions (use read-only-explorer).
Skill Claude CodeCodex
Spawn 3 adversarial subagents (Skeptic, Pragmatist, Critic) to evaluate a decision, architecture choice, or plan. Anti-anchoring: each role receives ONLY the question, not conversation history. Aggregates votes into consensus recommendation with dissent tracking. Use when facing critical decisions, architecture…
Skill Claude CodeCodex
Subagent/team delegation workflow. When NOT to use: pure read-only investigation (use read-only-explorer); pre-planning analysis (use scrutinize).
Skill Claude CodeCodex
Distill a person's (or field's) thinking into a runnable pi skill — research, extract, validate, generate. REQUIRED — read the full skill file first (multi-phase protocol with machine-checked gates); run the validate-run script on before claiming done — ALL-GREEN required. When NOT to use: software engineering…
Skill Claude CodeCodex
Distill software engineering expertise — an engineer's judgment, a codebase's conventions, or a domain's practice — and APPLY it to a TARGET project (never outputs a reusable skill file). REQUIRED — read the full skill file first (multi-phase protocol with machine-checked gates); run the validate-run script on before…
Skill Claude CodeCodex
Structured event logging for worker lifecycle, live agents, crash recovery. When NOT to use: high-level metrics dashboards (use observability-reliability); for runtime state inspection (use runtime-state-reader).
Skill Claude CodeCodex
Commit and release hygiene for safe version-control work. When NOT to use: experimental branches (use worktree-isolation); force-push or history rewrites (use scrutinize first).
Skill Claude CodeCodex
Iterative multi-round codebase audit with diminishing-returns detection. Run 5-20+ rounds, each focusing on one specific area. Built from 19 rounds of dogfooding pi-crew on itself. When NOT to use: one-shot review (use multi-perspective-review); investigation with single round (use read-only-explorer).
Skill Claude CodeCodex
Live agent registration, workspace isolation, termination, and eviction workflow. When NOT to use: static extension registration patterns (use pi-extension-lifecycle); ownership disputes (use ownership-session-security).
Skill Claude CodeCodex
Interactive waiting-task and mailbox workflow. When NOT to use: fire-and-forget background tasks; non-interactive worker contact.
Skill Claude CodeCodex
Model routing, parent context, thinking level, and prompt construction workflow. Use when changing model fallback, child Pi args, inherited context, task prompts, or compaction/compact-string behavior. When NOT to use: project-level config (use resource-discovery-config); routing for non-pi-crew agents.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: