thkt
192 verified skills1,886 total stars
use-context-reviewer-readability
Code readability review. Do NOT use for security (use-context-reviewer-security), error handling (use-context-reviewer-silence), or test design (use-context-reviewer-testability).
development11
use-cli-scout
Web search, page fetch, and GitHub repo exploration via scout CLI.
tools11
use-context-root-cause-analysis
Root cause analysis with 5 Whys.
research11
use-workflow-code
Workflow orchestration for /code.
development11
shake
Detect flaky tests by shaking them — repeated runs under varied order, parallelism, and seed — plus a static smell scan that flags latent flakiness in tests that currently pass. Classify each target as confirmed-flaky, latent-flaky, or stable and fix the root cause without weakening the test. Do NOT use to fix a confirmed single bug (use /fix) or for static-only code review (use /audit).
development11
shake
Detect flaky tests by shaking them — repeated runs under varied order, parallelism, and seed — plus a static smell scan that flags latent flakiness in tests that currently pass. Classify each target as confirmed-flaky, latent-flaky, or stable and fix the root cause without weakening the test. Do NOT use to fix a confirmed single bug (use /fix) or for static-only code review (use /audit).
development11
census
Discover design decisions that exist in code but have no DR, and produce a DR promotion candidate list ranked by impact and reversibility. Pairs with adrift, which scans existing DRs for drift against code.
development11
use-workflow-tdd-cycle
TDD with RGRC cycle and Baby Steps.
tools11
audit
Orchestrate reviewer agents under adversarial challenge (critic-audit + critic-evidence). Findings are positions to be argued, not facts to be aggregated. Do NOT use for quick PR screening (use /preview instead).
testing11
scribe
Extract recurring patterns from past closed PRs/issues and the research findings in workspace/research/, verify them against the latest code, and propose them to docs/wiki/ via PR.
development11
use-cli-herdr
Delegate implementation to codex (coder) via the herdr-agentchat plugin and drive a two-pane conversation to completion.
tools11
scribe
Extract recurring patterns from past closed PRs/issues and the research findings in workspace/research/, verify them against the latest code, and propose them to docs/wiki/ via PR.
development11
use-workflow-pageshot
Internal helper for /pr. Reads PR body (Preview URL + How to Test), drives agent-browser to capture a screenshot (1 step) or a video (2+ steps), and returns the artifact path for manual attachment on GitHub web UI.
tools11
use-cli-herdr
Delegate implementation to codex (coder) via the herdr-agentchat plugin and drive a two-pane conversation to completion.
tools11
adrift
Scan ADR Decision Outcome sections against current code and report drift with modification direction and priority. Do NOT use for repos without ADRs (use census instead).
development11
research
Probe project and technical questions. Findings are positions to be challenged with explicit sources, not conclusions. Phase 6 advisor pass argues against the synthesis before it lands. Do NOT use for design planning or plan generation (use /think instead).
testing11
use-context-reviewer-silence
Silent failure detection. Do NOT use for security (use-context-reviewer-security) or readability (use-context-reviewer-readability).
testing11
use-context-reviewer-testability
Testable code design review. Do NOT use for security (use-context-reviewer-security) or readability (use-context-reviewer-readability).
development11
use-workflow-pageshot
Internal helper for /pr. Reads PR body (Preview URL + How to Test), drives agent-browser to capture a screenshot (1 step) or a video (2+ steps), and returns the artifact path for manual attachment on GitHub web UI.
tools11
think
Design exploration with adversarial critique by critic-design. Assembles the surviving approach into a structured plan, self-checks it, and returns it to the caller. The issue's Plan section is the plan's only persistent home. Do NOT use for codebase investigation without planning intent (use /research instead).
development11
think
Design exploration with adversarial critique by critic-design. Assembles the surviving approach into a structured plan, self-checks it, and returns it to the caller. The issue's Plan section is the plan's only persistent home. Do NOT use for codebase investigation without planning intent (use /research instead).
development11
use-context-reviewer-readability
Code readability review. Do NOT use for security (use-context-reviewer-security), error handling (use-context-reviewer-silence), or test design (use-context-reviewer-testability).
development11
use-context-reviewer-silence
Silent failure detection. Do NOT use for security (use-context-reviewer-security) or readability (use-context-reviewer-readability).
testing11
use-context-reviewer-testability
Testable code design review. Do NOT use for security (use-context-reviewer-security) or readability (use-context-reviewer-readability).
development11
use-context-root-cause-analysis
Root cause analysis with 5 Whys.
research11
use-workflow-tdd-cycle
TDD with RGRC cycle and Baby Steps.
tools11
use-context-reviewer-security
OWASP Top 10 security review. Do NOT use for readability (use-context-reviewer-readability) or test design (use-context-reviewer-testability).
testing11
use-context-reviewer-security
OWASP Top 10 security review. Do NOT use for readability (use-context-reviewer-readability) or test design (use-context-reviewer-testability).
testing11
use-workflow-plan-preview
Internal helper for /think Step 11. Renders SOW.md + Spec.md as an integrated Astro view and returns a dev server URL.
tools11
adrift
Scan ADR Decision Outcome sections against current code and report drift with modification direction and priority. Do NOT use for repos without ADRs (use census instead).
development11
use-workflow-plan-preview
Internal helper for /think Step 11. Renders SOW.md + Spec.md as an integrated Astro view and returns a dev server URL.
tools11
checkout
Analyze Git changes and create a new branch with an appropriate name.
testing11
polish
Codex review + cleanup. Findings are challenged by critic-audit, not aggregated as facts, and fixes are applied directly. Do NOT use for internal multi-reviewer deep audits or findings reports (use /audit).
development11
use-workflow-code
Workflow orchestration for /code.
development11
fix
Rapidly fix small bugs and minor improvements in development environment. Do NOT use for new feature implementation or large-scale changes (refine via /issue and hand to the build workflow instead).
development11
commit
Analyze Git diff, generate a Conventional Commits format message, and run the commit.
development11
polish
Codex review + cleanup. Findings are challenged by critic-audit, not aggregated as facts, and fixes are applied directly. Do NOT use for internal multi-reviewer deep audits or findings reports (use /audit).
development11
challenge
Two-phase challenge that judges whether a discovered problem is real and whether a proposed idea is usable. Phase 1 loops subagent verification and advisor judgment over evidence (OUTCOME.md + parallel subagents) to self-resolve design-tree branches, asking the user only the irreversible residual and proceeding on stated assumptions for the rest. Phase 2 spawns two critic-design subagents (internal attack / OUTCOME.md attack) as devil's advocate input. The verdict leads the output as a simple GO / NO-GO. Do NOT use for code review findings (use /audit) or outcome assertion (use /assert which has built-in adversarial testing).
development11
checkout
Analyze Git changes and create a new branch with an appropriate name.
testing11
issue
Generate GitHub Issue with structured title and body. Standalone; requires no upstream stage. When challenge / research artifacts exist in the conversation, they feed the body's evidence; when a /think plan draft exists, it is transferred into the `## Plan` section.
research11
research
Probe project and technical questions. Findings are positions to be challenged with explicit sources, not conclusions. Phase 6 advisor pass argues against the synthesis before it lands. Do NOT use for design planning or plan generation (use /think instead).
testing11
challenge
Two-phase challenge that judges whether a discovered problem is real and whether a proposed idea is usable. Phase 1 loops subagent verification and advisor judgment over evidence (OUTCOME.md + parallel subagents) to self-resolve design-tree branches, asking the user only the irreversible residual and proceeding on stated assumptions for the rest. Phase 2 spawns two critic-design subagents (internal attack / OUTCOME.md attack) as devil's advocate input. The verdict leads the output as a simple GO / NO-GO. Do NOT use for code review findings (use /audit) or outcome assertion (use /assert which has built-in adversarial testing).
development11
outcome
Generates and updates .claude/OUTCOME.md interactively. When the file is absent or empty (no Behavior / all sections TBD), collects content via AskUserQuestion and writes the stub; when present, shows the current state and applies updates.
documentation11
use-cli-recall
Search past Claude Code/Codex sessions via recall CLI.
tools11
pr
Analyze branch changes and create a PR. Refine the body via prose review before posting.
tools11
slice
Break a plan / spec / PRD into independently-grabbable tracer-bullet vertical-slice issues and publish them to GitHub in dependency order. Each issue is one thin slice cutting through every layer.
tools11
dr
Create Decision Records (DR) in MADR v4 format with auto-numbering.
development11
pr
Analyze branch changes and create a PR. Refine the body via prose review before posting.
tools11
preview
AI screening review for PRs - preliminary check before human review. Do NOT use for deep multi-reviewer code quality audits (use /audit instead).
development11
census
Discover design decisions that exist in code but have no DR, and produce a DR promotion candidate list ranked by impact and reversibility. Pairs with adrift, which scans existing DRs for drift against code.
development11
use-cli-scout
Web search, page fetch, and GitHub repo exploration via scout CLI.
tools11
issue
Generate GitHub Issue with structured title and body. Standalone; requires no upstream stage. When challenge / research artifacts exist in the conversation, they feed the body's evidence; when a /think plan draft exists, it is transferred into the `## Plan` section.
research11
dr
Create Decision Records (DR) in MADR v4 format with auto-numbering.
development11
use-cli-recall
Search past Claude Code/Codex sessions via recall CLI.
tools11
use-cli-codegraph
Symbol-level code structure queries via codegraph CLI. Callers, callees, change impact, and symbol source trails.
tools11
fix
Rapidly fix small bugs and minor improvements in development environment. Do NOT use for new feature implementation or large-scale changes (refine via /issue and hand to the build workflow instead).
development11
skills/build
--- name: build description: Comprehensive implementation orchestrator chaining challenge / checkout / research / think / code / quality / commit / pr. Runs end-to-end from GO/NO-GO premise verification through branch creation, design, TDD implementation, a quality layer stacking internal audit and external polish (Codex + cleanup), commit, and PR creation. Use for any work spanning planning to implementation: new features, refactoring, migrations. Do NOT use for a single planned implementation
development11
code
Implement code following TDD/RGRC cycle with real-time test feedback. Vertical slices only (one test then one impl, never bulk tests then bulk impl). Do NOT use for small bug fixes or error resolution (use /fix instead).
development11
use-cli-codegraph
Symbol-level code structure queries via codegraph CLI. Callers, callees, change impact, and symbol source trails.
tools11
preview
AI screening review for PRs - preliminary check before human review. Do NOT use for deep multi-reviewer code quality audits (use /audit instead).
development11
slice
Break a plan / spec / PRD into independently-grabbable tracer-bullet vertical-slice issues and publish them to GitHub in dependency order. Each issue is one thin slice cutting through every layer.
tools11
outcome
Generates and updates .claude/OUTCOME.md interactively. When the file is absent or empty (no Behavior / all sections TBD), collects content via AskUserQuestion and writes the stub; when present, shows the current state and applies updates.
documentation11
commit
Analyze Git diff, generate a Conventional Commits format message, and run the commit.
development11
adrift
Scan ADR Decision Outcome sections against current code and report drift with modification direction and priority. Do NOT use for repos without ADRs (use census instead).
development11
assert
Independent outcome-based assertion with Codex + audit reviewers. Emits ternary Ready / Ready (caveat) / NotReady gate from reconciled static + dynamic evidence. Do NOT use for quick code review (use /polish) or static-only audit (use /audit).
development11
assert
Independent outcome-based assertion with Codex + audit reviewers. Emits ternary Ready / Ready (caveat) / NotReady gate from reconciled static + dynamic evidence. Do NOT use for quick code review (use /polish) or static-only audit (use /audit).
development11
adrift
Scan ADR Decision Outcome sections against current code and report drift with modification direction and priority. Do NOT use for repos without ADRs (use census instead).
development11
audit
Orchestrate reviewer agents under adversarial challenge (critic-audit + critic-evidence). Findings are positions to be argued, not facts to be aggregated. Do NOT use for quick PR screening (use /preview instead).
testing11
skills/build
--- name: build description: Comprehensive implementation orchestrator chaining challenge / checkout / research / think / code / quality / commit / pr. Runs end-to-end from GO/NO-GO premise verification through branch creation, design, TDD implementation, a quality layer stacking internal audit and external polish (Codex + cleanup), commit, and PR creation. Use for any work spanning planning to implementation: new features, refactoring, migrations. Do NOT use for a single planned implementation
development11
code
Implement code following TDD/RGRC cycle with real-time test feedback. Vertical slices only (one test then one impl, never bulk tests then bulk impl). Do NOT use for small bug fixes or error resolution (use /fix instead).
development11
probe
Extract repository spec while detecting bugs, spec gaps, and consistency drift via dual-purpose documentation. OUTCOME.md-axis question-driven exploration with ephemeral output. Do NOT use for code review (use /audit or /polish), feature implementation (use /code), planning only (use /think), or single-bug fix (use /fix).
development10
audit
Orchestrate reviewer agents under adversarial challenge (critic-audit + critic-evidence). Findings are positions to be argued, not facts to be aggregated. Do NOT use for quick PR screening (use /preview instead).
testing10
polish
Codex review + cleanup. Findings are challenged by critic-audit, not aggregated as facts, and fixes are applied directly. Do NOT use for internal multi-reviewer deep audits or findings reports (use /audit).
development10
adr
Create Architecture Decision Records (ADR) in MADR v4 format with auto-numbering.
development10
audit
Orchestrate reviewer agents under adversarial challenge (critic-audit + critic-evidence). Findings are positions to be argued, not facts to be aggregated. Do NOT use for quick PR screening (use /preview instead).
testing10
code
Implement code following TDD/RGRC cycle with real-time test feedback. Vertical slices only (one test then one impl, never bulk tests then bulk impl). Do NOT use for small bug fixes or error resolution (use /fix instead).
development10
probe
Extract repository spec while detecting bugs, spec gaps, and consistency drift via dual-purpose documentation. OUTCOME.md-axis question-driven exploration with ephemeral output. Do NOT use for code review (use /audit or /polish), feature implementation (use /code), planning only (use /think), or single-bug fix (use /fix).
development10
reflect
Surface the current session's tentative reflection (realization / judgment / counterfactual) and optionally promote it to the durable knowledge store. Escape hatch for when Stop hook reflection extraction failed.
data-ai10
code
Implement code following TDD/RGRC cycle with real-time test feedback. Vertical slices only (one test then one impl, never bulk tests then bulk impl). Do NOT use for small bug fixes or error resolution (use /fix instead).
development10
prototype
Build a throwaway prototype that answers a question. A runnable terminal app for logic/state questions, or several UI variants on one route for appearance questions. User-invoked only.
development10
shake
Detect flaky tests by shaking them — repeated runs under varied order, parallelism, and seed — plus a static smell scan that flags latent flakiness in tests that currently pass. Classify each target as confirmed-flaky, latent-flaky, or stable and fix the root cause without weakening the test. Do NOT use to fix a confirmed single bug (use /fix) or for static-only code review (use /audit).
development10
shake
Detect flaky tests by shaking them — repeated runs under varied order, parallelism, and seed — plus a static smell scan that flags latent flakiness in tests that currently pass. Classify each target as confirmed-flaky, latent-flaky, or stable and fix the root cause without weakening the test. Do NOT use to fix a confirmed single bug (use /fix) or for static-only code review (use /audit).
development10
prototype
Build a throwaway prototype that answers a question. A runnable terminal app for logic/state questions, or several UI variants on one route for appearance questions. User-invoked only.
development10
assert
Independent outcome-based assertion with Codex + audit reviewers. Emits ternary Ready / Ready (caveat) / NotReady gate from reconciled static + dynamic evidence. Do NOT use for quick code review (use /polish) or static-only audit (use /audit).
development10
reflect
Surface the current session's tentative reflection (realization / judgment / counterfactual) and optionally promote it to the durable knowledge store. Escape hatch for when Stop hook reflection extraction failed.
data-ai10
adr
Create Architecture Decision Records (ADR) in MADR v4 format with auto-numbering.
development10
assert
Independent outcome-based assertion with Codex + audit reviewers. Emits ternary Ready / Ready (caveat) / NotReady gate from reconciled static + dynamic evidence. Do NOT use for quick code review (use /polish) or static-only audit (use /audit).
development10
polish
Codex review + cleanup. Findings are challenged by critic-audit, not aggregated as facts, and fixes are applied directly. Do NOT use for internal multi-reviewer deep audits or findings reports (use /audit).
development10
use-context-reviewer-strictness
TypeScript type safety review. Do NOT use for testability (use-context-reviewer-testability), security (use-context-reviewer-security), or readability (use-context-reviewer-readability).
development9
use-cli-heptabase
Read and write Heptabase cards, journals, notes, tags, and AI Tutor data via heptabase CLI. All output is JSON.
tools9
skills/yomu-search
--- name: yomu-search description: Semantic code search (concept/identifier/related) for TS/JSX/CSS/HTML/Rust/Markdown via yomu CLI. Use when: concept search, identifier lookup, related code, 意味検索, 概念検索, hooks that do Y, where does X happen, unknown identifier. allowed-tools: [Bash, Read] user-invocable: false --- # yomu-search ## Commands | Purpose | Command | | ----------------- | ---------------------------------------------------
tools9
skills/root-cause-analysis
--- name: root-cause-analysis description: Root cause analysis with 5 Whys. Use when: root cause, 5 Whys, なぜなぜ分析, 根本原因, 原因分析, symptom fix, 対症療法. allowed-tools: [Read, Grep, Glob, Task] context: fork user-invocable: false --- # Root Cause Analysis - 5 Whys ## Principle Fix the root cause, not the symptom. Symptom fixes add complexity; root-cause fixes prevent recurrence. ## 5 Whys Process Ask "why" five times, descending through abstraction levels. | Step | Level | | ----
tools9
scout-search
Web search, page fetch, and GitHub repo exploration via scout CLI. Use when: web search, page fetch, deep research, GitHub repo exploration, latest docs, release notes, library docs, external info, WebFetch alternative, WebSearch alternative, 最新ドキュメント, リリースノート, 外部情報.
tools9
scout-search
Web search, page fetch, and GitHub repo exploration via scout CLI. Use when: web search, page fetch, deep research, GitHub repo exploration, latest docs, release notes, library docs, external info, WebFetch alternative, WebSearch alternative, 最新ドキュメント, リリースノート, 外部情報.
tools9
skills/recall-search
--- name: recall-search description: Search past Claude Code/Codex sessions via recall CLI. Use when: 前に, あの時, また同じ, あの件, past decisions, recurring mistake, module first contact, BACKLOG task pickup, temporal reference, structural echo, vague back-reference. allowed-tools: [Bash, Read] user-invocable: false --- # recall-search ## Triggers (call without deliberation) | Trigger | Signal | | -------------------- | ----------------------------------
tools9
skills/recall-search
--- name: recall-search description: Search past Claude Code/Codex sessions via recall CLI. Use when: 前に, あの時, また同じ, あの件, past decisions, recurring mistake, module first contact, BACKLOG task pickup, temporal reference, structural echo, vague back-reference. allowed-tools: [Bash, Read] user-invocable: false --- # recall-search ## Triggers (call without deliberation) | Trigger | Signal | | -------------------- | ----------------------------------
tools9
skills/google-workspace
--- name: google-workspace description: Access Google Sheets and Docs via gcloud CLI. Use when: Google Sheets/Docs URL, スプレッドシート, Sheets, Docs, Google ドキュメント. allowed-tools: [Bash, Read] user-invocable: false --- # Google Workspace ## Commands | Type | URL Pattern | Default | Structured | | ------ | --------------------------------- | ----------------------- | ----------------------------------------------- | | Sheet
tools9
skills/google-workspace
--- name: google-workspace description: Access Google Sheets and Docs via gcloud CLI. Use when: Google Sheets/Docs URL, スプレッドシート, Sheets, Docs, Google ドキュメント. allowed-tools: [Bash, Read] user-invocable: false --- # Google Workspace ## Commands | Type | URL Pattern | Default | Structured | | ------ | --------------------------------- | ----------------------- | ----------------------------------------------- | | Sheet
tools9
swarm
Large-scale parallel implementation with multi-agent swarm. Architect + QA + Implementer(s) collaborate via peer DM.
testing9
tailwind-patterns
Tailwind CSS v4 design decisions. Use when: Tailwind, utility-first, container query, design tokens, dark mode strategy.
tools9
use-cli-yomu
Semantic code search (concept/identifier/related) for TS/JSX/CSS/HTML/Rust/Markdown via yomu CLI.
tools9
explore
Explore codebase structure via iterative SQL queries against yomu's indexed SQLite DB. Do NOT use for concept/semantic search (use use-cli-yomu) or literal string search (use ugrep).
tools9
tuning
Empirical prompt tuning loop. Dispatch fresh subagents against skill/agent/command prompts, collect ambiguities, patch one at a time.
data-ai9
feature
Comprehensive feature development with exploration, architecture, TDD, and quality gates. Do NOT use for implementation without planning (use /code), bug fixes (use /fix), or planning only without implementation (use /think).
development9
audit-adr-gaps
Discover undocumented design decisions and challenge each candidate via critic-design before promotion. Rank by impact and reversibility, produce ADR promotion candidates. Treat each candidate as a position arguing for ADR status, not a fact to be filed. Pairs with audit-adr-drift, which scans existing ADRs for drift against code.
development9
use-cli-gcloud
Access Google Sheets and Docs via gcloud CLI.
tools9
use-cli-gcloud
Access Google Sheets and Docs via gcloud CLI.
tools9
use-cli-gh
GitHub CLI (gh) guide.
tools9
use-cli-gh
GitHub CLI (gh) guide.
tools9
use-cli-git
Git CLI guide.
tools9
use-cli-git
Git CLI guide.
tools9
use-cli-heptabase
Read and write Heptabase cards, journals, notes, tags, and AI Tutor data via heptabase CLI. All output is JSON.
tools9
use-cli-npm
npm scripts and CHANGELOG tools.
tools9
use-context-reviewer-performance
Frontend performance optimization.
development9
use-workflow-spec-validation
SOW/Spec cross-document consistency validation.
documentation9
audit-adr-drift
Scan ADR Decision sections against current code and report drift with modification direction and priority. Do NOT use for repos without ADRs (use audit-adr-gaps instead).
development9
accessing-google-workspace
Access Google Sheets and Docs via gcloud CLI. Use when user shares a Google Sheets/Docs URL or mentions スプレッドシート, Sheets, Docs, Google ドキュメント.
tools9
analyzing-root-causes
Root cause analysis with 5 Whys. Use when: root cause, 5 Whys, なぜなぜ分析, 根本原因, 原因分析, symptom fix, 対症療法.
research9
applying-code-principles
Fundamental software principles - SOLID, DRY, Occam's Razor, Miller's Law, YAGNI, Readable Code. Use when: 原則, シンプル, 複雑, リファクタリング, 保守性, clean code, best practice.
development9
applying-frontend-patterns
Framework-agnostic frontend component design patterns. Use when: component patterns, Container/Presentational, フロントエンドパターン, コンポーネント設計.
development9
brush
Enrich a vague prompt by routing to the best skill or injecting relevant rules and constraints. Use when prompt is rough or missing perspective. Examples: "find bugs", "refactor this", "add feature X", プロンプト補完, ブラッシュアップ. Do NOT use when the prompt is already detailed.
development9
formatting-audits
Binary Ready/NotReady gate for SOW/Spec. Use when: 設計書レビュー, 実装可否判定, document gate.
development9
integrating-storybook
Storybook integration with spec.md. Auto-generates Stories from Component API. Use when: storybook, stories, props, argTypes, variants, CSF3.
development9
optimizing-performance
Frontend performance optimization. Use when: performance, Web Vitals, LCP, FID, CLS, パフォーマンス最適化, 速度改善, bundle size.
development9
pageshot
Web app browser verification with screenshots. Reads operation_check.md, executes each step via agent-browser, and generates a self-contained HTML report with Base64-embedded screenshots. Use when: ブラウザ動作確認, スクショレポート, browser verification, screenshot report, operation check. Do NOT use for automated E2E tests (use /code with playwright).
development9
reviewing-testability
Testable code design review. Use when: テスト容易性, モック, mock-friendly, DI. Do NOT use for type safety (reviewing-type-safety), security (reviewing-security), or readability (reviewing-readability).
development9
reviewing-type-safety
TypeScript type safety review. Use when: type safety, any, 型安全, 型カバレッジ, strict mode. Do NOT use for testability (reviewing-testability), security (reviewing-security), or readability (reviewing-readability).
development9
scout
Read-through scan wrapper. Invokes scouting-anomalies on a file or glob. Use when user mentions /scout, scout, 通読スキャン, read-through, 違和感拾い, 気になるとこ洗い出し. Do NOT use for full audit (use /audit) or PR screening (use /preview).
testing9
scouting-anomalies
File-by-file read-through diagnostic to surface anomalies that fixed checklists miss. Use when: 通読, 違和感, 読み通し, 気になるとこ, read-through, anomaly scout, file-by-file review. Do NOT use for full audit (use /audit), PR screening (use /preview), or readability review (use reviewing-readability).
testing9
screening-pr-review
PR screening review before human review. Use when: スクリーニングレビュー, PRレビュー支援, レビュー前チェック, PR screening, pre-review.
tools9
utilizing-cli-tools
CLI tools guide (git, gh, npm). Use when: git操作, gh コマンド, npm スクリプト, HEREDOC コミット, CHANGELOG.
tools9
kousei
Proofread and auto-fix Japanese text using textlint. Use when user mentions 校正して, kousei, textlint, 日本語チェック, 文章チェック.
development9
vercel-react
React/Next.js performance rules from Vercel Engineering. Triggers on: React components, Next.js pages, data fetching, bundle optimization.
development9
verify
Independent outcome-based verification with Codex + audit reviewers. Emits binary Ready/NotReady gate from reconciled static + dynamic evidence. Use when user mentions 検証して, verify, 独立検証, outcome verification, gate decision, trust score (legacy), adversarial testing. Do NOT use for quick code review (use /polish) or static-only audit (use /audit).
development9
watch
Start Discord auto-response loop. Use when user mentions watch, discord watch, auto-response.
tools9
swarm
Large-scale parallel implementation with multi-agent swarm. Architect + QA + Implementer(s) collaborate via peer DM.
testing9
audit-adr-drift
Scan ADR Decision sections against current code and report drift with modification direction and priority. Do NOT use for repos without ADRs (use audit-adr-gaps instead).
development9
use-cli-npm
npm scripts and CHANGELOG tools.
tools9
audit-skill-craft
Judge a SKILL.md against craft axes (single responsibility, description distinctiveness, imperative voice, verifiable completion, calibration, progressive disclosure) and apply the fixes the audit surfaces. Do not use for format-presence-only checks (use reviewer-prompt) or reproducibility loops (use /tuning).
development9
audit-skill-craft
Judge a SKILL.md against craft axes (single responsibility, description distinctiveness, imperative voice, verifiable completion, calibration, progressive disclosure) and apply the fixes the audit surfaces. Do not use for format-presence-only checks (use reviewer-prompt) or reproducibility loops (use /tuning).
development9
audit-undocumented
Discover undocumented design decisions and challenge each candidate via critic-design before promotion. Rank by impact and reversibility, produce ADR promotion candidates. Treat each candidate as a position arguing for ADR status, not a fact to be filed.
testing9
explore
Explore codebase structure via iterative SQL queries against yomu's indexed SQLite DB. Do NOT use for concept/semantic search (use use-cli-yomu) or literal string search (use ugrep).
tools9
audit-undocumented
Discover undocumented design decisions and challenge each candidate via critic-design before promotion. Rank by impact and reversibility, produce ADR promotion candidates. Treat each candidate as a position arguing for ADR status, not a fact to be filed.
testing9
scheduled-tasks/refactor-scout
--- name: refactor-scout description: Weekly code quality patrol: review target repositories and report refactoring candidates as GitHub issues --- Review repositories and report areas for improvement as GitHub issues. ## Target Repositories | Repository | Branch | |-----------|---------| | thkt/kagami | main | | thkt/recall | main | ## Procedure Execute the following for each repository: ### 1. Clone Repository ```bash gh repo clone {repo} /tmp/refactor-scout-{repo_name} -- --depth 1 --b
development9
reviewing-type-safety
TypeScript type safety review. Use when: type safety, any, 型安全, 型カバレッジ, strict mode. Do NOT use for testability (reviewing-testability), security (reviewing-security), or readability (reviewing-readability).
development9
reviewing-silent-failures
Silent failure detection. Use when: silent failure, empty catch, エラーハンドリング, 握りつぶし, swallowed error. Do NOT use for security (reviewing-security), readability (reviewing-readability), or type safety (reviewing-type-safety).
testing9
vercel-react
React/Next.js performance rules from Vercel Engineering. Triggers on: React components, Next.js pages, data fetching, bundle optimization.
development9
utilizing-cli-tools
CLI tools guide (git, gh, npm). Use when: git操作, gh コマンド, npm スクリプト, HEREDOC コミット, CHANGELOG.
tools9
validating-specs
SOW/Spec cross-document consistency validation. Use when: 整合性チェック, consistency check, spec validation, 仕様検証.
testing9
screening-pr-review
PR screening review before human review. Use when: スクリーニングレビュー, PRレビュー支援, レビュー前チェック, PR screening, pre-review.
tools9
scouting-anomalies
File-by-file read-through diagnostic to surface anomalies that fixed checklists miss. Use when: 通読, 違和感, 読み通し, 気になるとこ, read-through, anomaly scout, file-by-file review. Do NOT use for full audit (use /audit), PR screening (use /preview), or readability review (use reviewing-readability).
testing9
scout
Read-through scan wrapper. Invokes scouting-anomalies on a file or glob. Use when user mentions /scout, scout, 通読スキャン, read-through, 違和感拾い, 気になるとこ洗い出し. Do NOT use for full audit (use /audit) or PR screening (use /preview).
testing9
reviewing-silent-failures
Silent failure detection. Use when: silent failure, empty catch, エラーハンドリング, 握りつぶし, swallowed error. Do NOT use for security (reviewing-security), readability (reviewing-readability), or type safety (reviewing-type-safety).
testing9
pageshot
Web app browser verification with screenshots. Reads operation_check.md, executes each step via agent-browser, and generates a self-contained HTML report with Base64-embedded screenshots. Use when: ブラウザ動作確認, スクショレポート, browser verification, screenshot report, operation check. Do NOT use for automated E2E tests (use /code with playwright).
development9
optimizing-performance
Frontend performance optimization. Use when: performance, Web Vitals, LCP, FID, CLS, パフォーマンス最適化, 速度改善, bundle size.
development9
integrating-storybook
Storybook integration with spec.md. Auto-generates Stories from Component API. Use when: storybook, stories, props, argTypes, variants, CSF3.
development9
formatting-audits
Binary Ready/NotReady gate for SOW/Spec. Use when: 設計書レビュー, 実装可否判定, document gate.
development9
orchestrating-workflows
Workflow orchestration for /code, /fix, /audit. Use when: /code ワークフロー, /fix ワークフロー, quality gates, 品質ゲート, RGRC サイクル.
development9
creating-adrs
Structured ADR creation in MADR format. Use when: ADR, Architecture Decision, 決定記録, 技術選定, アーキテクチャ決定, deprecation, 非推奨化.
development9
skills/root-cause-analysis
--- name: root-cause-analysis description: Root cause analysis with 5 Whys. Use when: root cause, 5 Whys, なぜなぜ分析, 根本原因, 原因分析, symptom fix, 対症療法. allowed-tools: [Read, Grep, Glob, Task] context: fork user-invocable: false --- # Root Cause Analysis - 5 Whys ## Principle Fix the root cause, not the symptom. Symptom fixes add complexity; root-cause fixes prevent recurrence. ## 5 Whys Process Ask "why" five times, descending through abstraction levels. | Step | Level | | ----
tools9
reviewing-readability
Code readability review. Use when: 可読性, 明確, 命名, 変数名, 関数名, ネスト, 関数設計, コメント, 複雑, Miller's Law, ミラーの法則, 認知負荷, AI-generated, 過剰設計. Do NOT use for security (reviewing-security), type errors (reviewing-type-safety), error handling (reviewing-silent-failures), or test design (reviewing-testability).
development9
glossary
Extract ubiquitous language from Slack and generate domain glossary. Use when user mentions 用語集, glossary, ユビキタス言語, ドメイン用語, Slack から用語.
data-ai9
orchestrating-workflows
Workflow orchestration for /code, /fix, /audit. Use when: /code ワークフロー, /fix ワークフロー, quality gates, 品質ゲート, RGRC サイクル.
development9
generating-tdd-tests
TDD with RGRC cycle and Baby Steps. Use when: TDD, テスト駆動, Red-Green-Refactor, Baby Steps.
development9
skills/yomu-search
--- name: yomu-search description: Semantic code search (concept/identifier/related) for TS/JSX/CSS/HTML/Rust/Markdown via yomu CLI. Use when: concept search, identifier lookup, related code, 意味検索, 概念検索, hooks that do Y, where does X happen, unknown identifier. allowed-tools: [Bash, Read] user-invocable: false --- # yomu-search ## Commands | Purpose | Command | | ----------------- | ---------------------------------------------------
tools9
accessing-google-workspace
Access Google Sheets and Docs via gcloud CLI. Use when user shares a Google Sheets/Docs URL or mentions スプレッドシート, Sheets, Docs, Google ドキュメント.
tools9
scheduled-tasks/refactor-scout
--- name: refactor-scout description: Weekly code quality patrol: review target repositories and report refactoring candidates as GitHub issues --- Review repositories and report areas for improvement as GitHub issues. ## Target Repositories | Repository | Branch | |-----------|---------| | thkt/kagami | main | | thkt/recall | main | ## Procedure Execute the following for each repository: ### 1. Clone Repository ```bash gh repo clone {repo} /tmp/refactor-scout-{repo_name} -- --depth 1 --b
development9
analyzing-root-causes
Root cause analysis with 5 Whys. Use when: root cause, 5 Whys, なぜなぜ分析, 根本原因, 原因分析, symptom fix, 対症療法.
research9
applying-code-principles
Fundamental software principles - SOLID, DRY, Occam's Razor, Miller's Law, YAGNI, Readable Code. Use when: 原則, シンプル, 複雑, リファクタリング, 保守性, clean code, best practice.
development9
applying-frontend-patterns
Framework-agnostic frontend component design patterns. Use when: component patterns, Container/Presentational, フロントエンドパターン, コンポーネント設計.
development9
extracting-ubiquitous-language
Slack conversation harvesting and term extraction. Internal module for /glossary. Use when: ユビキタス言語, 用語集, glossary, ドメイン用語.
business9
brush
Enrich a vague prompt by routing to the best skill or injecting relevant rules and constraints. Use when prompt is rough or missing perspective. Examples: "find bugs", "refactor this", "add feature X", プロンプト補完, ブラッシュアップ. Do NOT use when the prompt is already detailed.
development9
glossary
Extract ubiquitous language from Slack and generate domain glossary. Use when user mentions 用語集, glossary, ユビキタス言語, ドメイン用語, Slack から用語.
data-ai9
reviewing-testability
Testable code design review. Use when: テスト容易性, モック, mock-friendly, DI. Do NOT use for type safety (reviewing-type-safety), security (reviewing-security), or readability (reviewing-readability).
development9
validating-specs
SOW/Spec cross-document consistency validation. Use when: 整合性チェック, consistency check, spec validation, 仕様検証.
testing9
reviewing-security
OWASP Top 10 security review. Use when: security, OWASP, XSS, SQL injection, セキュリティ, 脆弱性, cloud security, AWS, IAM, Terraform, クラウドセキュリティ, インフラ. Do NOT use for readability (reviewing-readability), type errors (reviewing-type-safety), or test design (reviewing-testability).
testing9
verify
Independent outcome-based verification with Codex + audit reviewers. Emits binary Ready/NotReady gate from reconciled static + dynamic evidence. Use when user mentions 検証して, verify, 独立検証, outcome verification, gate decision, trust score (legacy), adversarial testing. Do NOT use for quick code review (use /polish) or static-only audit (use /audit).
development9
watch
Start Discord auto-response loop. Use when user mentions watch, discord watch, auto-response.
tools9
creating-adrs
Structured ADR creation in MADR format. Use when: ADR, Architecture Decision, 決定記録, 技術選定, アーキテクチャ決定, deprecation, 非推奨化.
development9
extracting-ubiquitous-language
Slack conversation harvesting and term extraction. Internal module for /glossary. Use when: ユビキタス言語, 用語集, glossary, ドメイン用語.
business9
generating-tdd-tests
TDD with RGRC cycle and Baby Steps. Use when: TDD, テスト駆動, Red-Green-Refactor, Baby Steps.
development9
kousei
Proofread and auto-fix Japanese text using textlint. Use when user mentions 校正して, kousei, textlint, 日本語チェック, 文章チェック.
development9
reviewing-security
OWASP Top 10 security review. Use when: security, OWASP, XSS, SQL injection, セキュリティ, 脆弱性, cloud security, AWS, IAM, Terraform, クラウドセキュリティ, インフラ. Do NOT use for readability (reviewing-readability), type errors (reviewing-type-safety), or test design (reviewing-testability).
testing9
reviewing-readability
Code readability review. Use when: 可読性, 明確, 命名, 変数名, 関数名, ネスト, 関数設計, コメント, 複雑, Miller's Law, ミラーの法則, 認知負荷, AI-generated, 過剰設計. Do NOT use for security (reviewing-security), type errors (reviewing-type-safety), error handling (reviewing-silent-failures), or test design (reviewing-testability).
development9
tailwind-patterns
Tailwind CSS v4 design decisions. Use when: Tailwind, utility-first, container query, design tokens, dark mode strategy.
tools9
use-context-reviewer-strictness
TypeScript type safety review. Do NOT use for testability (use-context-reviewer-testability), security (use-context-reviewer-security), or readability (use-context-reviewer-readability).
development9
use-cli-yomu
Semantic code search (concept/identifier/related) for TS/JSX/CSS/HTML/Rust/Markdown via yomu CLI.
tools9
use-context-reviewer-performance
Frontend performance optimization.
development9
feature
Comprehensive feature development with exploration, architecture, TDD, and quality gates. Do NOT use for implementation without planning (use /code), bug fixes (use /fix), or planning only without implementation (use /think).
development9
use-workflow-spec-validation
SOW/Spec cross-document consistency validation.
documentation9
audit-adr-gaps
Discover undocumented design decisions and challenge each candidate via critic-design before promotion. Rank by impact and reversibility, produce ADR promotion candidates. Treat each candidate as a position arguing for ADR status, not a fact to be filed. Pairs with audit-adr-drift, which scans existing ADRs for drift against code.
development9
tuning
Empirical prompt tuning loop. Dispatch fresh subagents against skill/agent/command prompts, collect ambiguities, patch one at a time.
data-ai9