Every skillMọi skill
Every hs:* skill is prose guidance — a structured prompt that tells Claude Code what to do. Invoke any with /hs:<name>. This is the exhaustive list, generated straight from each skill's SKILL.md; the curated catalog groups the most-used ones with how to pick.Mỗi skill hs:* là hướng dẫn prose — một prompt có cấu trúc cho Claude Code biết phải làm gì. Gọi bằng /hs:<tên>. Đây là danh sách đầy đủ, sinh trực tiếp từ SKILL.md của mỗi skill; catalog tuyển chọn nhóm các skill hay dùng kèm cách chọn.
SDLC coreLõi SDLC (13)
hs:code-review
Review code with technical rigor — bugs, regressions, security. Supports pending changes, PR number, commit hash, codebase scan. Verdict in review-decision.json; failing review blocks pr/ship/deploy.
hs:cook
Execute an approved plan phase by phase — TDD red→green, generate verification/review-decision artifacts, trace every step.
hs:debug
Systematic debugging with root-cause analysis before fixing. Use for bugs, test failures, unexpected behavior, performance issues, log analysis, CI/CD, and system investigation.
hs:fix
Fix bugs, test failures, and CI/CD failures with an evidence-based workflow. Use when there is a concrete bug, a clear error, or a red test.
hs:git
Git operations with conventional commits. Use for commit, push, PR, merge. Auto-split commits by type/scope. Scan for secrets before committing.
hs:plan
Create a verified implementation plan — research, constraint-scan, phase design, red-team, and validate before cook.
hs:review-pr
Review GitHub PR — diff, CI, correctness, security, breaking changes, anti-slop. Supports --fix (fix + commit) and --reply (post to GitHub). Use for a full PR review; for small local diffs use hs:code-review.
hs:scout
Fast codebase exploration using parallel agents — find files, locate code, gather context before implementing or debugging. Output goes to plans/reports/.
hs:setup
Configure this project's harness settings — terminal voice, guard/stage policy, output language — through the validated config CLIs. Use when onboarding a fresh install or changing a configuration decision.
hs:ship
Gated ship pipeline: review PASS → verification PASS → human approval → push/pr. Highest-risk stage — all gates must be real. Use when a completed branch needs to become a PR.
hs:test
Run and validate tests for the current change — unit/integration profiles, concise QA report, 100% pass gate.
hs:triage
Orchestrate the defect lifecycle — reproduce, classify, and gate bugs via hs:scout→hs:debug→hs:fix→hs:test. Use when a bug, test failure, or unexpected behavior is reported.
hs:understand
Orchestrate codebase comprehension before touching code — chain hs:repomix, hs:scout, hs:context-engineering to build a codebase map. Use before hs:plan on unfamiliar areas.
Flow & autonomyLuồng & tự chủ (11)
hs:afk
Run a plan/PRD in unattended (AFK) mode — preflight readiness, then route to Ralph sandbox or native fallback; loop commits freely in the middle, human reviews at both ends.
hs:autonomous-bell
Use when starting an unattended /goal or AFK loop that should end itself. Arms a reminder cron carrying a self-contained post-check protocol plus a consecutive-empty counter whose STOP is a deterministic off-decision read from disk — not a recollection — backed by the cron's 7-day expiry and session exit as hard floors.
hs:compound
Compound the harness's own self-knowledge — telemetry lenses + skill-formalization candidates + open backlog + a completeness critic — into ONE ranked self-improvement proposal, carried across runs in a durable belief store. Use when closing a work cycle to decide what to improve next. Advisory only; never edits skills/code/config; maintains only its own append-only belief store; never auto-creates skills.
hs:goal
Prepare a built-in /goal run at authoring time — interview the objective into a NEW self-contained goal.md, arm the autonomy bell, scaffold the intra-run cycle dir — then hand the loop to built-in /goal. Use when launching an unattended objective-until-met run that must survive being memory-blind between ticks.
hs:loop
In-session self-optimization loop — N iterations against a measurable metric, learns from git history, auto-keep/discard changes. Use for mechanical metrics (coverage, bundle size). In-session only.
hs:plans-kanban
View a file-based kanban board for all plans in plans/ — grouped by status (pending / in-progress / completed), navigate into a plan, check progress. Use for a quick work-status snapshot.
hs:project-management
Track plan progress, update task status, manage tasks, generate reports, coordinate docs updates. Use for project oversight, status checks, task hydration, and session handoffs.
hs:team
Orchestrate parallel Agent Teams — research, cook, review, debug with multiple independent teammates.
hs:vibe
One command takes a GitHub issue or feature request through the whole SDLC spine — worktree, planned gates, cook or fix, code-review, ship PR, review-pr, and optional merge with post-merge CI convergence. Use for issue→merged-PR autonomous runs.
hs:workflow-orchestrate
Design a spawn strategy for a delegated task — pick subagents vs Workflow vs Agent Teams, size and group the fan-out, batch-consolidate, early-write findings — then present it for approval before spawning. The delegation planner other skills call.
hs:worktree
Create, inspect, and clean up isolated git worktrees. Use for feature isolation, worktree health audits, and stale metadata cleanup.
Thinking & analysisTư duy & phân tích (7)
hs:bakeoff
Empirical bake-off — run 2-4 candidate probes on one mechanical metric, pick the winner by numbers or hand to a human when inside the noise band. Use when reasoning can't separate measurable options.
hs:brainstorm
Brainstorm solutions with honest trade-off analysis — ideation, architecture decisions, technical debate, feasibility exploration. Use before committing to an approach.
hs:critique
Multi-lens adversarial critique — fan independent lenses at an artifact, consolidate into one ranked verdict. Advisory by default; gate mode writes a machine verdict; optional --loop refute cycle.
hs:predict
5 expert personas independently debate a proposed change before implementation, catching architectural, security, performance, and UX risks early. Use before large features or high-risk changes.
hs:problem-solving
Structured unblocking — identify the block type, choose the right technique, reframe before resuming implementation. Use when stuck in a loop, complexity escalates, or a hypothesis fails 3+ times.
hs:scenario
Decompose a feature/code path across 13 dimensions to generate edge cases, risks, and test targets before implementation. Use for QA planning, risk discovery, and coverage audits.
hs:sequential-thinking
Multi-step analysis with revision — decompose complex problems, verify hypotheses, adjust direction mid-stream. Use when scope is unclear, reasoning needs a trace, or errors must be caught early.
ResearchNghiên cứu (7)
hs:autoresearch
Router for the autonomous iteration framework — routes to the right specialized skill (hs:loop, hs:plan, hs:discover) by goal. Use when unsure which skill in the autonomous-iteration group to invoke.
hs:contract-test
Validate a live contract by running the real probe — call the API/CLI and check status, exit code, and payload — then anchor the result as manual evidence. Use to prove a contract holds after building an endpoint or command, before ship. Never gate-driven.
hs:discover
Shape an ambiguous problem into a discovery brief for hs:plan — research + brainstorm chain -> direction summary, trade-offs, open questions.
hs:repomix
Pack a codebase or subtree into an AI-friendly digest (XML, Markdown, plain, JSON). Use for LLM context snapshots, external library audits, or large review prep. Pairs with hs:scout.
hs:research
Verified technical research — pose a question, gather multiple sources, verify evidence, synthesize a report. Use before implementing or when evaluating technology or architecture.
hs:security-scan
Scan codebase for security issues — hardcoded secrets, dependency CVEs, injection/authz gaps, STRIDE+OWASP. Use before releases, after adding auth/payment/data, or for periodic audits.
hs:techstack
Detect the target repo's tech stack (languages, test command, package manager, CI) so the harness adapts instead of assuming pytest. Use when the harness is installed into a non-Python repo, or before hs:test/hs:setup so commands match the real stack. Read-only; never mutates.
Memory & knowledgeBộ nhớ & tri thức (10)
hs:docs
Analyze codebase and manage project documentation — init, update, summarize. Use when docs need to be created, refreshed, or audited.
hs:docs-build
Đổ docs (md+frontmatter) + _index (graph+bands.yaml) + _present (presentation) → public/ (showcase HTML giữ theme/hiệu ứng) + dist/ (flat-md, pdf, excel). Nội dung sống ở md vật lý. Legacy showcase.yaml gộp = back-compat.
hs:docs-scaffold
Sinh skeleton doc từ template (frontmatter + heading + > TBD) theo required-set capability-driven. No-clobber. Không sáng tạo nội dung.
hs:docs-seeker
Look up library/framework documentation via llms.txt (context7.com) — API docs, GitHub repo analysis, latest features. Use when current official docs for a library are needed.
hs:docs-standardize
Validate structural docs (frontmatter + graph invariant) → artifact JSON → gate. Tìm gap vs template. KHÔNG sửa nội dung, KHÔNG phán chất lượng.
hs:document-skills
Create, edit, and analyze office files (.docx, .pdf, .pptx, .xlsx). Use to read content, create new files, edit with tracked changes, fill forms, or extract data from Word, PDF, PowerPoint, or Excel.
hs:insights
Surface read-only usage insights from harness telemetry — hot vs never-used skills, workflow chains, gate-block patterns — and propose end-user optimizations. Use when reviewing how the harness is actually used. Advisory only; never mutates.
hs:journal
Write a technical journal entry — record decisions, failures, and lessons learned after each session. Use after hs:ship, hs:cook, hs:fix, or when an incident needs honest documentation.
hs:remember
Capture the session's real knowledge — decisions made, non-obvious facts learned, user feedback — into the right durable home (DEC ledger, auto-memory, or BACKLOG) after you approve each entry. Use when you just made an architectural decision, learned something non-obvious, got correcting feedback worth keeping, or when the decision-capture nudge fires.
hs:retro
Generate data-driven retrospective reports from git history — velocity, code health, hotspots, plan progress. Use for sprint reviews, periodic assessments, or engineering lookbacks.
Create & generateTạo & sinh (6)
hs:agentize
Convert a codebase, feature, or module into an AI-agent-friendly npm CLI and/or MCP server. Use to expose existing capabilities as a reusable CLI tool or MCP server.
hs:bootstrap
Bootstrap a new project from scratch -- clarify requirements, init git, create doc structure, delegate to hs:plan + hs:cook. Use when starting a new repo or resetting a project's SDLC foundation.
hs:harness-creator
Create new harness primitives (hook, rule, schema, data, script, agent) — not skills. Use to add a hook/script/rule/schema/agent to the harness. To create an hs:* skill, use hs:skill-creator.
hs:mcp-builder
Build MCP servers for LLM external-service integration — FastMCP (Python), MCP SDK (Node/TypeScript), tool design, API integration, resource providers. Use to create or extend an MCP server.
hs:port
Extract, compare, port, or adapt a feature from a GitHub repository or local repo path into the current project. Use when copying behavior from another repo, studying how another codebase implements something, comparing implementations, or rewriting a feature in the local stack. Triggers on: 'port from', 'copy from repo', 'like how X does it', 'clone feature from', 'adapt from', 'borrow from', 'take from repo'.
hs:skill-creator
Create or update hs:* skills for the harness — SKILL.md, frontmatter, thin-core, references, validate via catalog.py. Use to create a new skill, refine triggers, or extend the harness.
Meta (harness authoring)Meta (tự dựng harness) (6)
hs:context-engineering
Manage the context budget — check limits, optimize tokens, coordinate subagents, debug context failures. Use when context is nearly full, designing multi-agent, or deciding what to load or drop.
hs:find-skills
Locate and route to the correct hs:* skill — analyze intent, query the hs plugin registry, return the exact invoke command. Use when unsure which skill fits or to browse the full hs:* catalog.
hs:project-organization
Organize files, directories, and content structure in any project. Use when creating files, determining output paths, organizing assets, or standardizing project layout.
hs:rule-author
Author a safe per-repo review-rule override in the layer-b folder docs/standards/ — detect conflicts with the shipped standards, narrow an overlapping scope when the intent differs, and on a real conflict confirm with you and record a mandatory reason. Use when you want to tune, disable, or add a review rule for this repo without editing the shipped standards tree, or when a review keeps flagging a rule that does not fit here.
hs:use
Run an install-disabled (off) skill from its stash + report; delegate discovery to hs:find-skills. Use when a skill is off.
hs:voice
Switch the terminal voice for this session — persona, harshness, explanation depth, no-markdown. Use when you want to change how the assistant talks (blunter, terser, a persona), or to persist a new default. (The code_style audience axis lives in output.yaml, set via output_config, not here.)
QualityChất lượng (1)
hs:manual-test
Run a session-based manual/exploratory test (SBTM charter → session → debrief) and emit anchored, admissible evidence. Use for behavior a static result file cannot capture — exploratory passes, manual API/UX checks, staging smoke.
SecurityBảo mật (1)
hs:cti-expert
Analyze cyber threat intelligence and OSINT cases. Use for exposure reviews, domain recon, breach checks, username/email/phone research, image forensics, blockchain tracing, darknet checks, cloud tenant recon, vulnerability lookup, threat modeling, and structured reports.
FrontendFrontend (11)
hs:design
Design brand identity, logos, banners, and visual assets. Use for brand systems, design tokens, corporate identity programs. Not for UI code patterns.
hs:frontend-design
Create polished frontend interfaces from designs/screenshots/videos. Use for web components, 3D experiences, replicating UI designs, quick prototypes, immersive interfaces, avoiding AI slop.
hs:frontend-development
Build React/TypeScript frontends with modern patterns. Use for components, Suspense, lazy loading, useSuspenseQuery, MUI v7 styling, TanStack Router, performance optimization.
hs:html-video
Create local MP4 videos from HTML/CSS/JS templates with nexu-io/html-video. Covers source checkout setup, template discovery, studio customization, preview, and render verification.
hs:react-best-practices
Apply React and Next.js performance optimization patterns from Vercel Engineering. Use for component optimization, rendering performance, bundle analysis.
hs:remotion
Build video content with Remotion in React. Use for programmatic video creation, animated sequences, data-driven video rendering.
hs:shader
Write GLSL fragment shaders for procedural graphics. Topics: shapes (SDF), patterns, noise (Perlin/simplex/cellular), fBm, colors (HSB/RGB), matrices, gradients, animations. Use for generative art, textures, visual effects, WebGL, Three.js shaders.
hs:stitch
AI design generation with Google Stitch. Generate UI designs from text prompts, export Tailwind/HTML/DESIGN.md, orchestrate design-to-code pipeline. Use for rapid prototyping, UI generation, design exploration.
hs:threejs
Build 3D web experiences with Three.js. Use for WebGL/WebGPU scenes, GLTF models, animations, physics, VR/XR. Supports 556 searchable examples.
hs:ui-styling
Style UIs with shadcn/ui components (Radix UI + Tailwind CSS). Use for accessible components, themes, dark mode, responsive layouts, design systems, color customization.
hs:ui-ux
UI/UX design intelligence for web and mobile: style selection, color systems, typography, layout, accessibility, interaction states, responsive behavior, forms, charts, design systems, and code review across React, Next.js, Vue, Svelte, SwiftUI, React Native, Flutter, Tailwind, shadcn/ui, and HTML/CSS. Use when designing, styling, or reviewing UI/UX for web or mobile.
BackendBackend (3)
hs:backend-development
Build backends with Node.js, Python, Go (NestJS, FastAPI, Django). Use for REST/GraphQL/gRPC APIs, auth (OAuth, JWT), databases, microservices, security (OWASP), Docker/K8s.
hs:better-auth
Add authentication with Better Auth (TypeScript). Use for email/password, OAuth providers (Google, GitHub), 2FA/MFA, passkeys/WebAuthn, sessions, RBAC, rate limiting.
hs:payment-integration
Integrate payments with SePay (VietQR), Polar, Stripe, Paddle (MoR subscriptions), Creem.io (licensing). Checkout, webhooks, subscriptions, QR codes, multi-provider orders.
FrameworksFramework (4)
hs:mobile-development
Build mobile apps with React Native, Flutter, Swift/SwiftUI, Kotlin/Jetpack Compose. Use for iOS/Android, mobile UX, performance optimization, offline-first, app store deployment.
hs:shopify
Build Shopify apps, extensions, and themes with Shopify CLI, GraphQL Admin API, Polaris UI, Liquid, webhooks, billing, and app configuration workflows.
hs:tanstack
Build with TanStack Start (full-stack React framework), TanStack Form (headless form management), and TanStack AI (AI streaming/chat). Use when creating TanStack projects, routes, server functions, forms, validation, or AI chat features.
hs:web-frameworks
Build with Next.js (App Router, RSC, SSR, ISR), Turborepo monorepos. Use for React apps, server rendering, build optimization, caching strategies, shared dependencies.
DatabaseCơ sở dữ liệu (1)
hs:databases
Design schemas, write queries for MongoDB and PostgreSQL. Use for database design, SQL/NoSQL queries, aggregation pipelines, indexes, migrations, replication, performance optimization, psql CLI.
AI / MLAI / ML (3)
hs:ai-artist
Generate product mockups, marketing assets, brand visuals, and concept art via Nano Banana with 129 curated prompts. Mandatory validation interview refines style/mood/colors (use --skip to bypass). 3 modes: search, creative, wild. Styles: Ukiyo-e, Bento grid, cyberpunk, cinematic, vintage patent.
hs:ai-multimodal
Analyze images/audio/video with Gemini API (better vision than Claude). Generate images (Imagen 4, Nano Banana 2, MiniMax), videos (Veo 3, Hailuo), speech (MiniMax TTS), music (MiniMax). Use for vision analysis, transcription, OCR, design extraction, multimodal AI.
hs:google-adk-python
Build AI agents with Google ADK Python. Multi-agent systems, A2A protocol, MCP tools, workflow agents, state/memory, callbacks/plugins, Vertex AI deployment, evaluation.
Dev toolsCông cụ dev (7)
hs:agent-browser
Automate browsers and apps with agent-browser. Use for testing, screenshots, forms, scraping, Browserbase/cloud browsers, and Electron when real Chrome cookies are not required.
hs:chrome-profile
Target a real Google Chrome profile for browser automation through Chrome DevTools MCP. Provides the chrome-profile CLI, profile discovery, live DevTools probing guidance, setup playbooks, and URL-anchor tab selection.
hs:gkg
Semantic code analysis with GitLab Knowledge Graph. Use for go-to-definition, find-usages, impact analysis, architecture visualization. Supports Ruby, Java, Kotlin, Python, TypeScript/JavaScript.
hs:llms
Generate llms.txt files from docs or codebase scanning. Follows llmstxt.org spec. Use for LLM-friendly site indexes, documentation summaries, AI context optimization.
hs:mintlify
Build and maintain Mintlify documentation sites. Covers docs.json, MDX components, navigation, page frontmatter, theming, OpenAPI/AsyncAPI, AI docs assets such as llms.txt and skill.md, deployment targets, and local validation CLI commands.
hs:use-mcp
Discover and execute MCP server tools. Two execution paths: LLM CLI (gemini + GEMINI_API_KEY primary, agy fallback) or direct scripts (deterministic, specific tool/server). Use for MCP integrations, tool execution, capability discovery, persistent tool catalog.
hs:web-testing
Web testing with Playwright, Vitest, k6. E2E/unit/integration/load/security/visual/a11y testing. Use for test automation, flakiness, Core Web Vitals, mobile gestures, cross-browser.
DevOpsDevOps (2)
hs:cleanup
Remove files an over-install left behind, safely: auto-clear pristine version-dropped files, and decide modified ones interactively (keep or change). Use after upgrading the harness or when verify reports orphans.
hs:release
Cut a harness release end to end: cut version, push tag, sync the public showcase, pack dist/ + gh release. Owner-only build step; pauses for confirmation before every outward-facing action. Use when cutting a new harness version.
InfrastructureHạ tầng (2)
hs:deploy
Deploy projects to any platform with auto-detection. Use when user says "deploy", "publish", "ship", "go live", "push to production", "host this app", or mentions any hosting platform (Vercel, Netlify, Cloudflare, Railway, Fly.io, Render, Heroku, TOSE, Github Pages, AWS, GCP, Digital Ocean, Vultr, Coolify, Dokploy). Auto-detects deployment target from config files and docs/deployment.md.
hs:devops
Deploy to Cloudflare (Workers, R2, D1), Docker, GCP (Cloud Run, GKE), Kubernetes (kubectl, Helm). Use for serverless, containers, CI/CD, GitOps, security audit.
VisualizationTrực quan hoá (6)
hs:drawio
Use when the user needs polished, editable diagrams (.drawio XML) with 10,000+ branded shapes (AWS/Azure/GCP/Cisco/Kubernetes/UML/ER/network), swimlanes, strict geometry, or export to PNG/SVG/PDF. Generates .drawio files that open in draw.io desktop and diagrams.net.
hs:excalidraw
Generate editable `.excalidraw` JSON files on canvas (architecture, data flow, system design). Choose when the diagram still needs further editing or an editable source is required.
hs:graphify
Build a queryable knowledge graph from code, docs, and media — architecture analysis, cross-file relationship discovery, token-efficient navigation. Use before hs:plan on unfamiliar codebases.
hs:mermaidjs
Text-based Mermaid.js v11 diagrams INLINE in markdown -- flowchart, sequence, class, ER, state, gantt, architecture -- renders natively on GitHub/GitLab without a separate file.
hs:preview
Explain a change or architecture with a diagram when visuals are clearer than prose — flow/architecture, before-after, sequence; output to docs/ or plans/.
hs:tech-graph
Generate publish-grade static technical diagrams (SVG+PNG, 8 styles) — architecture, data flow, sequence, agent/memory, concept map — for embedding in docs or slides.
UI / UXUI / UX (1)
hs:web-design-guidelines
Review UI code for Web Interface Guidelines compliance. Use when asked to "review my UI", "check accessibility", "audit design", "review UX", or "check my site against best practices".
MultimediaĐa phương tiện (1)
hs:media-processing
Process media with FFmpeg (video/audio), ImageMagick (images), RMBG (AI background removal). Use for encoding, format conversion, filters, thumbnails, batch processing, HLS/DASH streaming.
UtilitiesTiện ích (5)
hs:ask
Answer technical and architectural questions with expert analysis. Use for design decisions, best practices evaluation, solution comparison.
hs:copywriting
Conversion copywriting formulas, headline templates, email copy patterns, landing page structures, CTA optimization, and writing style extraction. Activate for writing high-converting copy, crafting headlines, email campaigns, landing pages, or applying custom writing styles.
hs:ghpm
GitHub project management for humans and AI agents. Use for issue/task planning, Projects boards, handoff status, gh/API automation, and CI-driven work tracking.
hs:markdown-novel-viewer
View markdown files in a calm, book-like reader served via HTTP. Use for long-form content review — RFCs, runbooks, design docs, reports, specs, novels — anywhere you want a distraction-free reading mode in the browser.
hs:watzup
Generate short handoff reports from Git branches, remote refs, worktrees, unfinished plans, and roadmap docs. Surfaces priority-ranked next steps with checkbox progress and rationale. Use when the user asks what's in flight, wants progress/next steps, is in a fresh worktree or detached checkout, or needs end-of-session status.
OtherKhác (1)
hs:show-off
Create preference-aware self-contained HTML pages to showcase work. Use for demos, visual presentations, interactive showcases.