Loading...
Loading...
Found 131 Skills
Assess tests for an identified code or test change, feature, PR or commit. USE FOR: review changed behavior or regression tests, prove pre-fix/missing behavior is detected, revision-bound red-green evidence. Polyglot. DO NOT USE without an identified change: hypothetical mutations or general suite blind spots (test-gap-analysis). Also exclude per-test readiness/A-F grades (grade-tests), coverage (coverage-analysis), new suites (code-testing), or mutation engines.
Audit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure.
Qualify and compare red and blue solutions for an Astro pull request using Astro-specific merge and tie-break criteria. Use only for Factory's adversary purple team.
Independently implement and validate an alternative solution for an Astro pull request. Use only for Factory's adversary blue team.
Turn a window of merged PRs (date, PR-number floor, or last 7 days) into a manual QA scenarios report for human testers — P0/P1/P2 routes with where to click, what to verify, what can go wrong — as markdown + HTML under paths.qa, shipped as a docs PR via om-auto-create-pr. Use for "what should QA test this week".
Diagnose a failure from its symptom — reproduce, classify the broken invariant, route via the repo AGENTS.md Task Router, trace the smallest root cause, and on request land a minimal fix behind a regression test. Framework knowledge via om-framework-context. Use for "why does this fail", "debug this", "fix this bug", "regression", "napraw błąd".
Turn a real agent failure into one reproducible eval-harness case plus the smallest knowledge change — failing case first, one knowledge owner, semantic assertions, before/after evidence through the repo's configured harness commands. Use for "add harness case", "agent got this wrong", "extend the harness", "rozszerz harness".
Compares an app clone against the original: a feature parity score from the feature matrix (weighted by must, should, could) with the missing list in build order, plus a screenshot layout diff that ignores colour so a rebrand does not count against you. Two standard-library Python tools. Use when the user says "how close is my clone", "compare it to the original", "what's missing", "parity check", "diff the screens", "is it ready", or after /replica-test.
Review code phong cách SENIOR 10 năm — bốn mục security · performance · naming · logic, mỗi mục cho ĐIỂM/10 + LỖI NẶNG NHẤT + CÁCH SỬA, rồi một kết luận PASS-sang-bước-kế hay CẦN-SỬA. Mục logic tìm bug (null/rỗng/âm/overflow, off-by-one, race) và SINH một test-case tái hiện cho mỗi bug (đặt tên qc-*), các test đó auto-chạy qua hook tất định (0-token, không LLM). Gọi khi user nói "qc code", "review code", "soi code", "chấm điểm code", "kiểm tra chất lượng code", "/qc-code". KHÁC /orca-sec-scans (Trivy quét TĨNH — vuln/misconfig/secret) và /code-review built-in (review tổng quát): qc-code là format senior bốn-mục-chấm-điểm-verdict + sinh test tái hiện. Mặc định review DIFF hiện tại; chỉ định file khi cần.
Investigate, optimize, or review software performance using representative measurements, profiles, budgets, and regression evidence. Use when latency, throughput, resource use, loading, rendering, bundle cost, or scalability is a primary concern. Do not use for speculative micro-optimization without a symptom, target, or demonstrated bottleneck.
Design, implement, repair, or review automated tests for observable behavior, regressions, contracts, stateful boundaries, and flaky failures. Use when testing is a primary part of the request or when choosing test levels, isolation, fixtures, or reliability requires dedicated judgment. Do not use merely to run existing checks after an ordinary implementation.
End-to-end defect hunt and repair across a whole codebase - context and business-rule capture, hierarchical sweeps (business logic, error handling, security, correctness, concurrency, data integrity, performance, config, test quality), reproduction, root-cause fixes, regression proofing, and an evidence-backed audit report. Use whenever bugs are to be found or fixed at repository or module scope, such as "find all the bugs", "fix everything", "audit this codebase", "review the whole codebase", "QA review", "harden this before release", "check for vulnerabilities", "why is this flaky". Use it equally for business-logic defects where nothing crashes but the result is wrong - "the calculation is wrong", "the numbers do not add up", "the discount applies twice", "a step can be skipped", "the request can be replayed", "the limit does not hold", "a user can see another user's data". Prefer this over ad-hoc reading whenever the scope is a repo, a module, or "everything" rather than one known failing line, and when reviewing AI-generated or recently merged code before shipping. This skill CHANGES CODE. If the user wants findings only with nothing modified - "review only", "do not touch anything", "report only" - use anti-bug-review instead.