Claude Skills by boshu2

Optionally analyze collections of durable verdicts for recurring evidence after the critical path. Triggers: "learn from verdicts", "mine validation history".

meta_skill (ms) — the skill-search/load engine over both corpora (agentops + jsm). Find a skill for a task, search skills, or load runnable skill guidance. Triggers: "ms", "meta_skill", "skill search", "find a skill for", "load skill guidance".

Use NTM as an optional pane adapter for caller-supplied roles and commands. Triggers: "ntm", "tmux panes", "ntm robot state".

Classify a pending decision as reversible or irreversible before it is acted on, and route irreversible ones to the caller instead of auto-deciding. Triggers: "is this a one-way door", "can we undo this", "should I just decide this", "the models disagree with me", before any auto-decided approval gate.

Distill repeated, evidence-backed expertise into a proposed skill, check, reference, or workflow artifact. Triggers: "operationalize this", "turn this expertise into a reusable capability".

Test repeated implementation shapes against independent exemplars and a holdout before routing an earned abstraction. Triggers: "mine a recurring code pattern", "is this abstraction earned", "extract invariants from implementations".

Shape or refine the existing bead or caller intent without a second planning artifact. Triggers: "plan", "discover and plan", "shape this goal".

Optionally test a retrospective causal question against durable verdict evidence. Triggers: "postmortem", "causal retrospective", "test a retrospective hypothesis".

Optionally challenge a frozen plan with one fresh independent judge before implementation. Triggers: "premortem", "challenge this plan", "what could make this plan fail".

Create or refine PRODUCT.md while separating evidence, aspiration, users, value, and non-goals. Triggers: "product", "create PRODUCT.md", "product boundary".

Use RCH once to offload a build or collect remote-compilation diagnostics. Triggers: "use RCH", "offload this build".

Compare a claimed state with observable repository evidence and report concrete gaps. Requires a claim or expected state to test. Triggers: "reality check", "is this claim actually done", "compare claim to repo".

Execute one behavior-preserving structural transformation and report evidence. Triggers: "refactor this", "simplify without changing behavior".

Answer a bounded question with current cited evidence. Triggers: "research", "investigate this question", "find evidence". (Investigating a repository routes to codebase-recon.)

Reverse-engineer an authorized repo, binary, or product into a verifiable feature inventory and adoption map. Triggers: "reverse-engineer X", "tear down Y", "what should we steal from Z", "evaluate competitor/upstream", "should we fork/adopt/build-native".

Pick the one AgentOps skill that owns a request, or answer that none does. Triggers: "which skill covers this", "route this", "is there a skill for X", "what should I use here", or any request whose owning skill is not obvious.

Coordinate one RPI traversal: one bounded Plan, Implement, and fresh Validate experiment, then report and stop. Triggers: "run rpi", "run one traversal", "execute this plan", orchestration or worker delegation that implements changes.

Inspect disk pressure with SBH and run one explicitly authorized recovery action. Triggers: "check disk pressure", "run SBH".

Stamp a bounded project, component, or CI scaffold and verify the generated result once. Triggers: "scaffold", "create project component or boilerplate".

Review the bead or caller intent write scope for completeness and ambiguity. Triggers: "review write scope", "check scope boundaries", "scope this change".

Run authorized repository security scans for vulnerabilities, dependency risk, secrets, and binary policy. Triggers: "security", "run repository security scans for", "security skill".

Retired — its runtime-neutrality contract moved to docs/contracts/runtime-neutrality.md. Triggers: none — not routable.

Create a metadata-complete AgentOps skill source package, regenerate its derived projections, and check or repair structural hygiene in skill packages. Triggers: "create a skill", "scaffold skill", "absorb external skill", "new skill", "heal skill", "repair skill hygiene", "audit skill structure", "check skill package".

Author and tier behavioral probes for a skill, including seeded-defect probes that escape ceiling saturation. Triggers: "measure this skill", "the probe came back INERT", "the control arm aces it", "harden this scenario", "is this skill actually doing anything".

Load only the standards relevant to a caller-supplied change, then report concrete findings. Triggers: "check standards", "which standards apply".

Report observable AgentOps evidence without selecting work. Triggers: "status", "show AgentOps status".

Dispatch explicit disjoint packets exactly once through a caller-selected executor. Triggers: "swarm", "dispatch disjoint packets", "parallel explicit tasks".

Generate tests and coverage plans. Triggers: "test", "generate tests and coverage plans.", "test skill".

Mine caller-supplied usage history for repeated toil and emit ranked evidence. Triggers: "mine toil", "find repeated operational work".

Operate the Agentic Coding Flywheel as a caller-selected software factory; keep its runtime state out of AgentOps verdicts. Triggers: "using flywheel", "agent flywheel".

Operate a caller-selected Gas City 1.4 with upstream registry packs and native run-centered surfaces while keeping GC runtime state out of AgentOps verdicts. Triggers: "using gc", "gas city", "drive the mayor", "dispatch through gc".

Freshly judge whether a finished change is actually proven against bead or caller acceptance — the independent verdict before merge; optionally persist verdict.v2 for a declared consumer, and stop. Triggers: "validate", "independently validate", "is this proven", "vibe".

Scaffold an explicit one-shot workflow adapter without lifecycle authority. Triggers: "build a workflow adapter", "scaffold a one-shot workflow".

Execute one bounded RED to GREEN experiment from bead or caller intent; return derived subject identity and check facts. Triggers: "implement", "implement this bead", "run the experiment". Full plan-to-validation requests route to rpi.

Execute one bounded RED to GREEN experiment from bead or caller intent; return derived subject identity and check facts. Triggers: "implement", "implement this bead", "run the experiment". Full plan-to-validation requests route to rpi.

Freshly judge exact subject content against bead or caller acceptance, optionally persist verdict.v2 for a declared consumer, and stop. Triggers: "validate", "independently validate", "vibe".

Operate a caller-selected Gas City 1.4 with upstream registry packs and native run-centered surfaces while keeping GC runtime state out of AgentOps verdicts. Triggers: "using gc", "gas city", "drive the mayor", "dispatch through gc".

Switch a caller-selected coding-agent Triggers: "switch account", "rotate coding-agent account".

Use Agent Mail as an optional messaging and Triggers: "coordinate writers", "reserve files".

Operate explicit orchestrator, implementer Triggers: "agent-native factory", "role-shaped agent panes", "persistent workers".

Use an explicitly selected AGY runtime for Triggers: "agy", "antigravity", "AGY evidence".

Guard outcome work against process overhead. Triggers: RPI pre-Plan guard; explicit "full anti-ceremony audit" requests.

Front door for agent automation: choose Triggers: "build automation", "which orchestration shape", "should this use NTM".

Mine past agent sessions for working Triggers: "cass", "mine past agent sessions for", "cass skill".

Configure default Claude Code enforcement Triggers: "cc-hooks", "configure Claude Code hooks", "install hooks".

Reconstruct a repository as cited Triggers: "codebase recon", "trace this codebase", "repository audit", "refresh the prior recon".

Run one caller-supplied Codex command Triggers: "run Codex headless", "capture Codex evidence".

Collect independent perspectives for an Triggers: "council", "multi-judge review", "independent perspectives".

Compile or lint a persistent Mayor-style Triggers: "craft a goal prompt", "mayor goal", "goal-runner prompt", "lint this goal", "is this goal safe". (Shaping one experiment''s intent routes to plan.)

Handle blocked destructive commands and Triggers: "dcg", "handle a DCG block", "configure agent safety guardrails".