Use when adding or changing tests, the e2e harness, fake CLIs in pipeline-step tests, test process isolation, or CI test sharding.

Use when changing fork routing, forge-profile identity, repository URL persistence, credential redaction, the repo-config trust boundary (trusted-default-branch-only fields), gate-agent instruction neutralization, or repository-declared extra gates.

Use when changing macOS release signing, release artifact verification, the release workflow, or the self-update channel manifest.

Use when changing subprocess execution, cancellation, process-tree cleanup, orphan reaping, or Windows child hardening.

Use when changing PR body rendering, home-path redaction, artifact path publication, pipeline-attestation markers, or pre-push attestation.

Use when changing or migrating the shared require-no-mistakes PR-enforcement action, its live PR lookup, or its workflow caller.

Use when changing review sessions, finding decisions, the review fix-round carry, review schema retries, agent timeouts, local Test behavior and its live-validation contract, approval overrides and parked gates, or intent conformance.

Use when changing documentation ownership, generated agent guidance, review auto-fix guidance, the combined document+lint housekeeping pass, or the agent memory-file prompt rules.

Use when changing daemon startup, singleton ownership, shutdown, logging, event subscriptions, lifecycle commands, or the daemon's login-shell environment.

Use when changing CI readiness, forge check collection, reruns, CI timeouts, CI step findings and fix rounds, or PR lifecycle monitoring.

Use when changing local branch synchronization, custody recovery, post-review head binding, rebasing or rebase.strategy, private-mirror reconciliation, or force-push safety.

Use when changing agent model or effort configuration, per-run Pi profile pins, adapter mappings, or eval candidate profiles.

Create JUI controls — interactive form field components used in GWT/JUI applications. Use this skill when the user asks to: create a new control, build a custom form field, create an interactive input component, or asks about control patterns. Also trigger when the user mentions control, form field, value management, dirty detection, or asks how to create a reusable input that can be placed in a ControlForm.

Convergence-paced review-resolve loop over a change and its governing surfaces. Verifies each finding against the codebase and the base it is measured from, then re-reviews until each is disposed of.

This skill should be used when choi names a core protocol whose Lean contract is to be re-derived, or invokes /rederive. Runs the per-protocol flow: gather the pre-understanding, derive the protocol's epistemic solution and fuse it with choi's horizon, then read the current contract against it, through sketch, edit, dogfood, review, merge, the chart close and a retrospective on this skill. Project-local contributor tooling.

Scoped empirical validation. Decomposes a target↔surrogate equivalence claim into facets, bounds a test space, captures evidence inside it, and carries the untested complement forward.

Deficit Recognition Probe — surface multiple deficit hypotheses for the user's current situation and route by user-constituted recognition (fit review, not protocol scoring).

Quick protocol recommendation from recent sessions, or quest-based learning through scenario, trial, and quiz.

Delegate factual research to a Codex CLI session that uses Codex's builtin `goal` command and invokes Aitesis (`$inquire`) for Tavily-based external verification. User-invoked via /goal-research.

Context about to be acted on may no longer hold — stale, weakly sourced, or contradicted: challenge each claim the action leans on, and the user judges what stands before acting.