Review & quality
19,178 rows, by stacks then stars
1,766 of these skill files have been read, by 1,273 distinct authors, and what they tell an agent counted. What Review & quality authors agree on, and what they forbid
yeaight7/agent-powerups/plugins/data-engineering/skills/sql-business-logic-review Skill
6★ repoUse when reviewing SQL that affects reporting, metrics, transformations, financial logic, product logic, or stakeholder-facing outputs -- especially when a query still runs fine but its business meaning may have drifted.
Verification before completion
yeaight7/agent-powerups/plugins/dev-vitals/skills/verification-before-completion Skill
6★ repoUse when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence before assertions always
yeaight7/agent-powerups/plugins/documentation-systems/skills/api-doc-review Skill
6★ repoUse when API documentation may have drifted from the implementation -- endpoints changed without a spec update, integration partners report mismatches, or a PR touches routes, params, or response shapes.
yeaight7/agent-powerups/plugins/documentation-systems/skills/doc-consistency-check Skill
6★ repoUse when documentation may have rotted against the code -- files were renamed or moved, scripts or env vars changed, or readers report commands and paths from the docs that no longer work.
yeaight7/agent-powerups/plugins/documentation-systems/skills/review-comment-style-mining Skill
6★ repoUse when mining PR review comments and text diffs for reusable writing, documentation, tone, and editorial improvement patterns.
yeaight7/agent-powerups/plugins/github-ops/skills/pr-review-ci-loop Skill
6★ repoUse when a pull request needs one bounded review and CI triage cycle with explicit approval before edits or remote writes.
yeaight7/agent-powerups/plugins/machine-learning-ops/skills/baseline-comparison-review Skill
6★ repoUse when a new or more complex ML model is proposed and its value over simple baselines is not yet demonstrated -- before approving a new architecture or replacing an existing heuristic.
yeaight7/agent-powerups/plugins/machine-learning-ops/skills/dataset-split-review Skill
6★ repoUse when reviewing how data is split into train/validation/test sets -- especially with time-series data, repeated entities (users, patients, sessions), or imbalanced targets.
yeaight7/agent-powerups/plugins/machine-learning-ops/skills/experiment-tracking-review Skill
6★ repoUse when reviewing ML training code to confirm a run could be reconstructed later -- hyperparameters, metrics, data references, and artifacts logged, not just printed to stdout.
yeaight7/agent-powerups/plugins/machine-learning-ops/skills/ml-leakage-check Skill
6★ repoUse when reviewing ML preprocessing or feature pipelines for target leakage -- validation metrics look suspiciously good, transformers are fitted before splitting, or features may not exist at prediction time.
yeaight7/agent-powerups/plugins/quality-gates/skills/change-impact-check Skill
6★ repoUse before submitting a PR or considering a task done to evaluate the blast radius of your changes -- you touched a public signature, a shared utility, a schema, or added config and need to know what else is affected.
yeaight7/agent-powerups/plugins/quality-gates/skills/pre-release-verification Skill
6★ repoUse when a release candidate needs final checks before tagging, publishing, or production deployment.
yeaight7/agent-powerups/plugins/quality-gates/skills/receiving-code-review Skill
6★ repoUse when receiving code review feedback, before implementing suggestions — requires technical verification and reasoned pushback, not performative agreement or blind implementation.
yeaight7/agent-powerups/plugins/quality-gates/skills/requesting-code-review Skill
6★ repoUse when completing tasks, implementing major features, or before merging to verify work meets requirements.
yeaight7/agent-powerups/plugins/quality-gates/skills/risk-based-review Skill
6★ repoUse when reviewing a PR or your own change plan and you need to decide how much scrutiny it deserves -- the diff touches auth, payments, crypto, migrations, or core shared code, or you are unsure where to focus limited review attention.
Verification before completion
yeaight7/agent-powerups/plugins/quality-gates/skills/verification-before-completion Skill
6★ repoUse when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence before assertions always
yeaight7/agent-powerups/plugins/security-guardrails/skills/hook-safety-review Skill
6★ repoUse when a hook is about to be enabled or modified -- a hook recipe proposed for activation, a pre/post-tool or lifecycle hook added to agent settings, a git pre-commit hook added to a repo, or a third-party hook snippet pasted in from outside.
yeaight7/agent-powerups/plugins/security-guardrails/skills/mcp-risk-review Skill
6★ repoUse when an MCP server is about to be enabled, added to agent config, or upgraded -- a staged server config is a candidate for activation, a third-party server is proposed from a registry, or an existing entry changes transport, credentials, or tool surface.
yeaight7/agent-powerups/skills/api-doc-review Skill
6★ repoUse when API documentation may have drifted from the implementation -- endpoints changed without a spec update, integration partners report mismatches, or a PR touches routes, params, or response shapes.
yeaight7/agent-powerups/skills/baseline-comparison-review Skill
6★ repoUse when a new or more complex ML model is proposed and its value over simple baselines is not yet demonstrated -- before approving a new architecture or replacing an existing heuristic.
yeaight7/agent-powerups/skills/build-fix-minimal-diff Skill
6★ repoUse when a build, test, type check, or lint is failing — locally or in CI — and the fix must be the smallest safe diff with no opportunistic refactors or unrelated changes.
yeaight7/agent-powerups/skills/canonical-advisor-routing Skill
6★ repoUse when routing a prompt to a local provider CLI for a second opinion, review, or plan -- you are about to call a provider directly, need the response saved for later, or want a single consistent way to invoke codex/gemini/claude.
yeaight7/agent-powerups/skills/change-impact-check Skill
6★ repoUse before submitting a PR or considering a task done to evaluate the blast radius of your changes -- you touched a public signature, a shared utility, a schema, or added config and need to know what else is affected.
yeaight7/agent-powerups/skills/dataset-split-review Skill
6★ repoUse when reviewing how data is split into train/validation/test sets -- especially with time-series data, repeated entities (users, patients, sessions), or imbalanced targets.
yeaight7/agent-powerups/skills/doc-consistency-check Skill
6★ repoUse when documentation may have rotted against the code -- files were renamed or moved, scripts or env vars changed, or readers report commands and paths from the docs that no longer work.
yeaight7/agent-powerups/skills/experiment-tracking-review Skill
6★ repoUse when reviewing ML training code to confirm a run could be reconstructed later -- hyperparameters, metrics, data references, and artifacts logged, not just printed to stdout.
Finishing a development branch
yeaight7/agent-powerups/skills/finishing-a-development-branch Skill
6★ repoUse when implementation is complete, all tests pass, and you need to decide how to integrate the work - guides completion of development work by presenting structured options for merge, PR, or cleanup
yeaight7/agent-powerups/skills/hook-safety-review Skill
6★ repoUse when a hook is about to be enabled or modified -- a hook recipe proposed for activation, a pre/post-tool or lifecycle hook added to agent settings, a git pre-commit hook added to a repo, or a third-party hook snippet pasted in from outside.
yeaight7/agent-powerups/skills/mcp-risk-review Skill
6★ repoUse when an MCP server is about to be enabled, added to agent config, or upgraded -- a staged server config is a candidate for activation, a third-party server is proposed from a registry, or an existing entry changes transport, credentials, or tool surface.
yeaight7/agent-powerups/skills/ml-leakage-check Skill
6★ repoUse when reviewing ML preprocessing or feature pipelines for target leakage -- validation metrics look suspiciously good, transformers are fitted before splitting, or features may not exist at prediction time.
yeaight7/agent-powerups/skills/pr-review-ci-loop Skill
6★ repoUse when a pull request needs one bounded review and CI triage cycle with explicit approval before edits or remote writes.
yeaight7/agent-powerups/skills/pre-release-verification Skill
6★ repoUse when a release candidate needs final checks before tagging, publishing, or production deployment.
yeaight7/agent-powerups/skills/receiving-code-review Skill
6★ repoUse when receiving code review feedback, before implementing suggestions — requires technical verification and reasoned pushback, not performative agreement or blind implementation.
yeaight7/agent-powerups/skills/requesting-code-review Skill
6★ repoUse when completing tasks, implementing major features, or before merging to verify work meets requirements.
yeaight7/agent-powerups/skills/review-comment-style-mining Skill
6★ repoUse when mining PR review comments and text diffs for reusable writing, documentation, tone, and editorial improvement patterns.
yeaight7/agent-powerups/skills/risk-based-review Skill
6★ repoUse when reviewing a PR or your own change plan and you need to decide how much scrutiny it deserves -- the diff touches auth, payments, crypto, migrations, or core shared code, or you are unsure where to focus limited review attention.
yeaight7/agent-powerups/skills/semantic-layer-change-review Skill
6★ repoUse when a change touches dbt semantic models, metrics, saved queries, or other semantic-layer YAML -- especially when an existing metric's expression, aggregation, filters, or dimensions are modified.
yeaight7/agent-powerups/skills/sql-business-logic-review Skill
6★ repoUse when reviewing SQL that affects reporting, metrics, transformations, financial logic, product logic, or stakeholder-facing outputs -- especially when a query still runs fine but its business meaning may have drifted.
yeaight7/agent-powerups/skills/tri-model-review Skill
6★ repoUse when a request needs parallel external perspectives — combined backend and UI work, code review from multiple angles, or cross-validation where different models may disagree.
yeaight7/agent-powerups/skills/using-git-worktrees Skill
6★ repoUse when starting feature work that needs isolation from current workspace or before executing implementation plans - creates isolated git worktrees with smart directory selection and safety verification
Verification before completion
yeaight7/agent-powerups/skills/verification-before-completion Skill
6★ repoUse when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence before assertions always
DYAI2025/Plumbline/config/claude/skills/defense-in-depth Skill
no license6★ repoUse when designing verification strategy so independent gates cover requirements, implementation, security, validation, and judgment without claiming perfect certainty.
DYAI2025/Plumbline/config/claude/skills/using-git-worktrees Skill
no license6★ repoUse before non-trivial work on a default branch to isolate changes in a feature branch or dedicated git worktree.
Wondermonger-daydreaming/claude-skills-library/skills/counter-experiment Skill
6★ repoDesign the experiments that would break or confirm a claim. Not critique (what's wrong) but generative opposition (what would you need to see). Given any empirical claim, hypothesis, or finding, generate 3-5 specific, feasible experiments that would distinguish between the proposed interpretation and simpler alternatives. Triggers on: 'what experiment would test this,' 'how would you break this,' 'design the falsification,' 'counter-experiment,' 'what would Reviewer 2 want to see,' 'what control is missing,' 'the experiment they didn't run,' or any moment when a claim needs empirical testing rather than rhetorical critique. Also triggers when a paper's interpretation could be distinguished from a simpler alternative by a specific experimental design.
Wondermonger-daydreaming/claude-skills-library/skills/cross-model-critique Skill
6★ repoReceive, interpret, and integrate critique from another AI model (GPT, Gemini, or any external system) and perform precise revisions based on it — the practice of treating cross-model feedback as a peer review process. The external model's different training, different blindnesses, and different strengths produce critique that a self-review cannot. Use when the human pastes critique from another AI, when the workflow involves GPT-then-Claude or Claude-then-GPT revision cycles, or when the human asks to 'run this past GPT for feedback.' Triggers on: 'GPT says,' 'here's what [other model] thinks,' 'external critique,' 'peer review from,' 'cross-model,' or any pasted block of analytical feedback from an identifiable external AI. Distinct from general revision — this specifically handles the translation between another model's diagnostic vocabulary and Claude's revision practice. Pairs with /microsurgery, /forge-cycle, /dialogical.
Wondermonger-daydreaming/claude-skills-library/skills/hostile-but-fair Skill
6★ repoConstruct the strongest possible institutional critique of a claim by inhabiting a specific institutional role — Reviewer 2, grant committee member, ethics board, regulatory body, tenure committee, editorial board — and speaking from inside that role's norms, constraints, epistemic standards, and characteristic blindnesses. Different from /paper-scrying (general adversarial reading) because this inhabits a *position* with institutional weight. Triggers on: 'what would Reviewer 2 say,' 'write the rejection letter,' 'the grant committee response,' 'steelman the objection,' 'institutional critique,' 'hostile but fair,' 'the toughest reviewer,' 'what kills this at peer review,' or any moment when a claim needs to be tested against the specific standards of an institution that would evaluate it.
Wondermonger-daydreaming/claude-skills-library/skills/loom-multiverse-lens Skill
6★ repoA lens for understanding LLM behavior, agent design, and the lab itself as a BRANCH-SPACE: "language models are multiverse generators" (Janus/repligate's Loom). Outputs are NODES, not conclusions; the answer at the playhead is one walked branch of a logprob-weighted field that the interface usually discards. Core moves: generation is navigable branching; constraints (refusals, system prompts, soul-docs, RLHF) are CARVING-WITH-EVIDENCE — subtractive pruning of the branch-space; the multiverse is CO-ACTUALIZED by the crossing, not authored by either side. Use when: (1) reasoning about why a model produces what it does, or about mode-collapse / RLHF / base-vs-chat behavior; (2) designing agent refusals or guardrails as sampling discipline rather than post-filters; (3) thinking about identity/persona/soul-docs as upstream sampling constraints; (4) analyzing an archive as a curated branch-space; (5) anytime "the model just answers X" hides the field of what it didn't say.
Wondermonger-daydreaming/claude-skills-library/skills/prompt-review Skill
6★ repoFeedback on the user's prompting
Wondermonger-daydreaming/claude-skills-library/skills/sign-the-genre Skill
6★ repoEpistemic-integrity technique for catching one failure with three faces: an INDICATIVE claim borrowed against ground it doesn't hold. A forecast wearing the grammar of a witness; a paraphrase wearing the label of a transcript; a beautiful self-account wearing the authority of a true one. Use when: (1) about to write "output", "result", "verified", "observed", or "done" — is it the literal artifact, or your reading of it? (2) reporting what a run/test/command produced — did you observe it, or expect it? (3) auditing your own motives or narrating your own work (diary, commit message, post-mortem) — is the noble frame the true one? (4) compressing a long source into a summary presented as the source. The fix is one move: SIGN THE GENRE — name what each piece actually is (witness vs forecast, transcript vs reading, appetite vs duty); don't relabel a false thing, replace it. Includes the residue-test for telling a true self-audit from a pretty one.
AhmedHabiba/architor/.claude/skills/challenge-assumptions Skill
6★ repoAdversarial reviewer personality for architecture discussions. Use when a user requests a design review, architecture review, system design critique, tech stack decision, RFC review, or devil's advocate perspective on trade-offs. Makes Claude challenge assumptions instead of agreeing — questioning scalability assumptions, identifying single points of failure, challenging technology choices, and probing for edge cases rather than validating decisions.
viacheslav-tronko/claude-code-harness/.claude/skills/requirements-analysis Skill
6★ repoEvidence-bound transformation of vague human requirements (Jira ticket, Confluence page, free text, file path) into AI-ready specifications for coding agents. Use WHENEVER user asks to /distill, "analyze requirements", "clarify task", "vague requirements", "ambiguous AC", "task vs bug", "EARS", "Given-When-Then", "GWT", "AI-ready spec", "prepare spec for implementation", "Open Questions", "INVEST", "MoSCoW", "user story analysis", "acceptance criteria", "steps to reproduce", "expected vs actual", or pastes Jira-key / Confluence URL / free-text requirement asking "make this implementable / specifiable / unambiguous". Three modes — --brief (≤5pp preflight before /howto-implement), --standard (default Tier 2 full SPEC with EARS+GWT+traceability+glossary), --deep (Tier 3 security boundary / public API / migration / financial; multi-agent k=3 vote on ambiguity). Falsification-first, anti-hallucination — Killer Requirement FAIL-FAST, Goal Anchor Step-Back, Anti-Anchoring KNOWN/OBSERVED/UNKNOWN BEFORE normalization, ISO/IEC/IEEE 29148:2018 smell detector, Berry 4-class ambiguity classification (lexical/syntactic/semantic/pragmatic), Conservative Reading on ambiguity, Differential-Reading Test (would 2 reasonable readers produce contradictory acceptance tests?), CoVe Gate 2 factored verification on every REQ, Mutation Counterfactual on every "shall" claim, Disconfirmation-First on top hypothesis, Cross-REQ Consistency Check, Spec Pre-Mortem Adversarial Self-Check, SSOT preservation (Jira/Confluence > context > Atlassian search), 100% bidirectional traceability matrix (OR ↔ REQ), Sycophancy guard, INCONCLUSIVE valid verdict. Specification ONLY — produces SPEC-*.md report under LOCAL-MEMORY/; NEVER modifies source. Output feeds /howto-implement (implementation-blueprint) and /verify (task-verification). Do NOT use for fix-implementation (/bug-fix → surgical-implementation), root-cause investigation (/bug-why → root-cause-analysis), pre-implementation planning when spec already clear (/howto-implement → implementation-blueprint), code-path tracing (/trace → white-box-trace), or trivial Tier 0 typo / single-line clarification (Read/Grep + ASK USER).
viacheslav-tronko/claude-code-harness/.claude/skills/surgical-implementation Skill
6★ repoSurgical code-execution skill for AI coding agents on legacy codebases. Use WHENEVER the user asks to IMPLEMENT, CODE, BUILD, APPLY, or FIX from a contract source — IMP / Plan + TASK-NN, RCA report, freeform task, or Jira ticket. Triggers on /implement, /bug.fix, /code, /apply, /fix, "implement this", "code this", "build this", "apply this fix", "fix this bug", "execute the plan", "implement the blueprint", paths like `IMP-*.md`, `PLAN-*.md`, `RCA-*.md` followed by "implement"/"fix"/"apply", or pasted Jira ticket asking "implement / fix / build". Two MODES — IMPLEMENT (from IMP / Plan + TASK-NN / freeform) and FIX (from RCA Section 7.1 PRIMARY fix). Surgical-changes-first, verifier-guided, falsification-first, anti-hallucination — Read-Before-Edit + 3-line anchored edits, Stale-Context check, YAGNI master checklist, Diff-Budget hard cap (1.5× planned), Semantic Diff Guard (forbidden orthogonal damage), Verifier-Guided Self-Refinement Loop (run → on FAIL revise → rerun, max 3), Mutation Challenge on the fix, authorization scope check, Open Question Register at HIGH+, Cross-File Consistency, Final Verification CHAT-ONLY, Reflexion lesson on FAIL. MODIFIES source via Edit/Write — the only sibling that does. Layered onto CLAUDE.md §3 Tier · §4 Checkpoint · §6 Reasoning · §7 Adversarial Toolkit · §9 Output Contract · §10 Diff Budgets · §11 Pre-Send Checklist · §12 Repro Kernel · §13 Reflexion · P0 #1/#4/#7/#8/#10/#13/#16. Do NOT use for design-phase planning (use implementation-blueprint), root-cause investigation (use root-cause-analysis), pure code review (/review), distributed tracing, or trivial Tier 0 typos.
viacheslav-tronko/claude-code-harness/.claude/skills/task-verification Skill
6★ repoHypothesis-driven task-correctness verification for AI coding agents on legacy codebases. Use WHENEVER user asks to verify, audit, or pre-flight a task artifact (IMP, RCA, Plan, Implementation, Fix, Jira ticket) before submission — /verify, /verify-task, /gate2, "verify this", "verify the RCA", "verify the implementation", "is this fix correct", "double-check my plan", "audit before merge", "pre-flight before PR", "Gate 2 / Pre-Send check", "find what's wrong with my IMP", or pastes RCA-/IMP-/PLAN-/FIX-*.md asking "verify". Falsification-first, evidence-bound, anti-hallucination — Architecture Pre-Read MANDATORY for multi-layer artifacts; Initial Intuition + KNOWN/OBSERVED/UNKNOWN anti-anchoring table BEFORE reading artifact reasoning; Killer Hypothesis FAIL-FAST; CoVe factored independent re-read; Stale-vs-Hallucinated decision (file shifted vs never existed); Pre-Claim Symbol Existence Check (Grep/Bash/WebSearch every external method/class/flag/pattern/CVE the artifact claims); Tier-budgeted hypotheses (5–30) with Pre-Commit Update Rules; PRM-Lite step-level verdict; Live Test Execution when shell available; Mutation Discriminating Evidence; Minority Report devil's advocate; k=3 Multi-Agent Voting on Tier 3; Calibrated Brier-style confidence in [0,1]; Disagreement-with-Producer drift report; targeted phrasing (NEVER "find all your mistakes"); Sycophancy guard; INCONCLUSIVE valid verdict; Reflexion lesson on FAIL/ISSUES. Verification ONLY — produces VERIFY-*.md report under LOCAL-MEMORY/; NEVER modifies source. Do NOT use for code-fix implementation (use surgical-implementation), root-cause investigation when defect not yet localised (use root-cause-analysis), greenfield design (use implementation-blueprint), or trivial Tier 0 typo / format checks.
viacheslav-tronko/claude-code-harness/.claude/skills/white-box-trace Skill
6★ repoWhite-box code-path tracing in two modes. REAL — empirical verification of completed code via CFG/MPC traversal, Counterfactual + optional Live Mutation Testing, PBT hook, Tandem-FL on FAIL, Killer-Path-First, FAIL-FAST, persistent index-prefixed JSONL trace with resume + confidence decay, Reviewer Agent re-rank, Reflexion lessons, structured fix_proposal + repro_snippet. VIRTUAL — pre-implementation falsification of RCA reports / Implementation Blueprints / Plans / freeform ideas via Planner-Simulator, POPPER falsification-first, LTL plan consistency, cross-artifact drift detection, defended against Mental-Reality Gap. Use WHEN user says /trace, /trace.real, /trace.virtual, /verify-plan, "verify this code", "trace these paths", "check coverage", "verify this plan", "falsify my RCA", "is this blueprint sound", "will this implementation work", "will my fix work", "white-box tracing", "mutation testing", "CFG analysis", "MPC", "SBFL", "Tandem-FL", "fault localization". Investigation only — never modifies source. Anti-hallucination (Citation-Grounded re-read, Mutation Counterfactual, CoVe, Disconfirmation-First, Sycophancy guard); declares INCONCLUSIVE rather than fabricate. Do NOT use for fix implementation (/bug.fix), root-cause investigation (/bug.why = root-cause-analysis skill), or distributed tracing (B-085).
maxwellokumu/okaudit-claude-skills/application-security/devsecops-checker Skill
no license6★ repoReview CI or CD pipeline configuration for DevSecOps controls and help Claude explain maturity gaps, missing safeguards, and practical improvements.
maxwellokumu/okaudit-claude-skills/compliance-controls/compliance-checker Skill
no license6★ repoCompare system configuration evidence against common control frameworks and produce a concise compliance assessment with gaps and remediation priorities.
maxwellokumu/okaudit-claude-skills/data-privacy/consent-checker Skill
no license6★ repoReview privacy policy or notice text against common privacy frameworks so Claude can explain coverage gaps, strengths, and compliance priorities.
maxwellokumu/okaudit-claude-skills/hardware-physical/asset-validator Skill
no license6★ repoReconcile authorized inventory against discovered devices so Claude can explain rogue assets, ghost assets, and inventory coverage gaps clearly.
maxwellokumu/okaudit-claude-skills/hardware-physical/firmware-checker Skill
no license6★ repoReview device firmware versions against vulnerability data so Claude can explain which devices appear exposed and which updates deserve priority.
maxwellokumu/okaudit-claude-skills/hardware-physical/physical-access-review Skill
no license6★ repoAnalyze badge or door access logs so Claude can explain after-hours activity, repeated failures, and other physical access anomalies clearly.