dotclaude

Use when auditing comments, docstrings, or embedded documentation prose in a scope — to strip bloat and rewrite cluttered-but-justified entries. Triggers on "audit comments", "clean up docstrings", "strip stale comments", "comment cleanup".

Use when the goal is to reduce code size, remove unnecessary complexity, or simplify a module without removing user-facing behavior.

Use when auditing and fixing end-to-end contracts and compatibility in a scope — HTTP/RPC APIs, events/messages, schemas, files, CLIs, libraries, configuration, database consumers, and external integrations. Triggers on "audit contracts", "API compatibility review", "schema compatibility audit", "check producer/consumer drift", "compatibility review", and "contract review". Performs an architecture-led review of semantic, backward, forward, and mixed-version compatibility; named interface classes are a minimum, never an exhaustive boundary.

Use when auditing and fixing logic bugs in a scope — off-by-one, wrong operator, inverted condition, unhandled boundary cases, unit/dimension mismatch, code that doesn't do what its name or docstring claims. Triggers on "audit correctness", "fix logic bugs", "check for off-by-one", "correctness review".

Use when auditing and fixing end-to-end data integrity in a scope — invariant violations, partial or conflicting writes, unsafe migrations/backfills, duplicate or reordered effects, precision or time corruption, stale derived stores, broken retention/deletion, and unrecoverable drift. Triggers on "audit data integrity", "review data consistency", "check migrations and invariants", "find data corruption", and related requests. Named risks are minimum, never exhaustive.

Use when auditing and fixing error handling in a scope — swallowed exceptions, defaults returned on error, optional chaining masking missing data, floating promises / missing awaits, entry points with no error boundary, resource leaks on error paths. Triggers on "audit error handling", "hunt silent failures", "fix swallowed errors", "check error boundaries", "missing await".

Use when auditing and fixing logging in a scope — missing observability, INFO bloat that doesn't scale, stale or wrong-level messages, payload dumps on routine paths. Triggers on "audit logs", "fix logging", "logging review", "clean up log noise".

Use when cutting code in a folder that doesn't earn its keep — whether a solution is over-engineered, where maintenance cost can be reduced, or where a feature should be removed entirely.

Use when auditing and unifying pattern consistency in a scope — divergent implementations of the same concern, reimplementations of existing utilities, naming violations, import disorder, unused dependencies. Triggers on "audit patterns", "fix inconsistencies", "unify conventions", "fix reimplementations", "pattern review".

Use when auditing and fixing performance hot paths in a scope — N+1 queries, overfetching, wasted recomputation, blocking hot paths, unbounded growth. Triggers on "audit performance", "fix slow paths", "check for N+1", "performance review".

Use when auditing and fixing reliability or resilience in a scope — timeouts, retries, idempotency, partial failure, dependency outages, overload, failover, recovery, reconciliation, and startup/shutdown behavior. Triggers on "audit reliability", "review resilience", "failure recovery review", "resilience audit", "check graceful degradation", and related failure-tolerance intent. Named failure classes are a minimum, never an exhaustive boundary.

Use when auditing and fixing security in a scope — injection (SQL, shell, template), auth/authz gaps, secrets in code or logs, weak crypto, missing validation at trust boundaries, XSS, SSRF, path traversal. Triggers on "audit security", "security review", "fix vulnerabilities", "OWASP check". Performs a comprehensive, architecture-led review; named weakness classes are a minimum, never an exhaustive boundary.

Use when running the full audit sweep on a scope — runs the focused audit dimensions (necessity, structure, patterns, correctness, error handling, logs, perf, security, tests, complexity, comments) as a parallel-worktree workflow that integrates in dependency order. Triggers on "audit", "full audit", "audit everything".

Use when auditing repo topology in a scope — code grouped by type where the repo groups by feature, god-modules, misplaced files, leaky or missing module boundaries, layer-direction violations, import cycles, over-nesting. Triggers on "audit structure", "repo structure review", "fix module boundaries", "check layering", "files in the wrong place".

Use when the user wants a thorough audit of the existing test suite — aggressively cutting tests that exist just to exist, rewriting weak ones to actually catch the bugs they should, identifying production code that is uncovered or thinly covered, and adding tests where coverage would catch real defects. Trigger on "audit tests", "find test bloat", "tests don't catch anything", or similar requests.

Use when the user asks to merge another branch into the current branch, pull upstream changes without rebasing, or resolve merge conflicts.

Use when the user asks to rebase the current branch onto another branch or update branch history without merging.

Produce an up-to-date, state-of-the-art recommendation using domain expertise and brief web research. Use only when explicitly invoked.

Use when the user wants to examine a plan, decision, belief, or problem through a rigorous Socratic inquiry before acting.

Use when a task needs an implementation-ready design document before code changes begin.

Use when the user asks to commit, create a git commit, checkpoint changes, save work to git, or write a commit message.

Use when the user wants a UI/UX/frontend *build request* rewritten into an expert design brief before an LLM builds from it. Rewrites the prompt, never executes it; sharpens vague gestures into named techniques, makes implicit constraints explicit, and surfaces absent load-bearing dimensions as author-facing fill slots — never fabricating specifics. Do NOT trigger for design exploration or alternative directions (use design), page/flow usability review (use ux), person-to-person voice rewrites (use writing-identity), or LLM prompt engineering (use prompt). Triggers - "refine this design prompt", "make this UI prompt better", "turn this into a design brief", or "/design-refiner" with a prompt or path.

Use when the requested deliverable is distinct competing approaches to compare before choosing what to implement.

Use when the user asks to create or update a pull request for the current branch.

Use when reviewing a user-facing page or flow for usability issues, confusion, friction, or missed UX principles.

Use when rewriting an existing skill so Opus 5 (or any frontier reasoning model) executes it with better judgment, less scaffolding, and tighter calibration. Do NOT trigger for new skills from scratch (use skill-creator) or for prompt artifacts (use prompt).

Use when the user wants an independent expert review of work done in this conversation before accepting or extending it.

Use when rewriting AI-sounding communication into natural language that sounds like its author, or when drafting in the user's voice, for Slack and chat messages, DMs, emails, technical or project updates, client notes, feedback, explanations, and similar person-to-person writing. Trigger on requests such as 'make this sound human', 'less AI', 'less polished', 'sound like me', or 'rewrite this naturally'. Do NOT use when another person's voice must be preserved, for generic marketing copy, or when the user wants only proofreading without a voice rewrite.

Use when the user asks to create, refine, evaluate, or optimize an LLM system prompt.

Use when the user needs a monthly work summary, time-tracking report, or commit-based activity recap.