Agent skills
Skills you can use with AI coding agents, indexed from public GitHub repositories.
-
stop-slop
Strip AI writing patterns from prose across five registers — casual, professional, academic, technical, and narrative. Applies 200+ pattern rules to eliminate predictable LLM tells. Triggers on clean up writing, remove AI slop, edit prose, make this sound less AI.
majiayu000/claude-skill-registry 163
-
Global Validation
Data validation strategies for stock_picker_bot using Pydantic models and SQLAlchemy constraints. Covers API input validation, stock data validation, and database integrity.
majiayu000/claude-skill-registry 163
-
competitor-intel-brief
Run a structured competitive teardown in 20 minutes. Covers positioning, ICP, offer analysis, moat assessment, vulnerability mapping, and a direct attack brief. Use when someone says "analyze my competitor," "competitive research," "how do I differentiate from X," or "what are my competitor's weaknesses."
majiayu000/claude-skill-registry 163
-
gemma_domain_trainer_prototype
Gemma Domain Trainer (Prototype)
majiayu000/claude-skill-registry 163
-
epf-dump
Разобрать EPF-файл обработки 1С (EPF/ERF) в XML-исходники. Используй когда пользователь просит разобрать, декомпилировать обработку, получить исходники из EPF/ERF файла
majiayu000/claude-skill-registry 163
-
pdf
Comprehensive PDF manipulation toolkit for extracting text and tables, creating new PDFs, merging/splitting documents, and handling forms. When Claude needs to fill in a PDF form or programmatically process, generate, or analyze PDF documents at scale.
majiayu000/claude-skill-registry 163
-
vercel-deployment
Expert knowledge for deploying to Vercel with Next.js Use when: vercel, deploy, deployment, hosting, production.
majiayu000/claude-skill-registry 163
-
aget-create-skill
Create new AGET skills following S-V-O naming and standard templates. Use when developing new /aget-* skills with proper vocabulary, specs, and traceability.
majiayu000/claude-skill-registry 163
-
bmad-teach-me-testing
Teach testing progressively through structured sessions. Use when user says "lets learn testing" or "I want to study test practices"
majiayu000/claude-skill-registry 163
-
mobile-games
Mobile game development principles. Touch input, battery, performance, app stores.
majiayu000/claude-skill-registry 163
-
rust-backend
Rust backend developer for building high-performance, memory-safe APIs and services. Use when building Rust backends with Axum, Actix-web, Rocket, or Warp. Covers Tokio async runtime, SQLx, Diesel, error handling, and zero-cost abstractions.
majiayu000/claude-skill-registry 163
-
gamma
Gamma sample skill for local discovery and CLI smoke tests.
majiayu000/claude-skill-registry 163
-
beads-state
Git-friendly, crash-recoverable state persistence for the Gastown orchestration chipset. Manages agent identities, work items, hooks, convoys, and merge requests as JSON files with atomic write operations.
majiayu000/claude-skill-registry 163
-
cloudflare-browser-rendering
Add headless Chrome automation with Puppeteer/Playwright on Cloudflare Workers. Use when: taking screenshots, generating PDFs, web scraping, crawling sites, browser automation, or troubleshooting XPath errors, browser timeouts, binding not passed errors, session limits, page.evaluate __name errors, or waitForSelector timeout issues.
majiayu000/claude-skill-registry 163
-
Plugin Forge
Step-by-step guidance for plugin forge.
majiayu000/claude-skill-registry 163
-
research
Deep codebase exploration. Triggers: research, explore, investigate, understand, deep dive, current state.
majiayu000/claude-skill-registry 163
-
Swarm Orchestration
Orchestrate multi-agent swarms with agentic-flow for parallel task execution, dynamic topology, and intelligent coordination. Use when scaling beyond single agents, implementing complex workflows, or building distributed AI systems.
majiayu000/claude-skill-registry 163
-
ollama
Run local LLMs with Ollama. Configure models, manage resources, and integrate with applications. Use for local AI, private LLM deployments, and offline inference.
majiayu000/claude-skill-registry 163
-
kb-mastodon
Background knowledge about Mastodon, the dominant federated microblogging platform and de facto reference implementation for ActivityPub social networking. Covers Mastodon's JSON-LD @context structure (three-element array: activitystreams, security/v1, inline object defining toot: http://joinmastodon.org/ns# with Emoji/discoverable/featured/featuredTags/ blurhash/votersCount/indexable/memorial/suspended/attributionDomains, schema: http://schema.org# with PropertyValue/value, as: with Hashtag/manuallyApprovesFollowers/movedTo/sensitive), content-type and Accept header requirements (application/activity+json or application/ ld+json with ActivityStreams profile, content negotiation for actor discovery), HTTP signature requirements (RSA-SHA256, Digest header mandatory for POST, signed headers include request-target/host/date/ digest), authorized fetch / secure mode (AUTHORIZED_FETCH env var, requires signatures on GET requests too, breaks unsigned federation), supported object types (Note and Question first-class, Article/Page/ Image/Audio/Video/Event converted with content or name + URL appended), poll implementation (Question wrapped in Create, oneOf for single-choice anyOf for multiple, options as Note with name and replies.totalItems, votes as Create of Note matching option name, endTime/closed/votersCount), edit/Update handling (since 3.5.0, requires updated timestamp, sends Update activity with full object, edit history not yet federated), Announce/boost behavior (object field is URI string, receiver fetches original, shares collection on statuses), direct message addressing (no as:Public, no follower collection, only actor URIs in to, all must be Mentioned in tag), visibility calculation (public=as:Public in to, unlisted=as:Public in cc, private=followers in to/cc without as:Public, limited=actors not all Mentioned, direct=all actors Mentioned), custom emoji (tag type Emoji with name shortcode and icon Image), HTML sanitization (allowed: a/p/br/span/b/i/em/strong/ blockquote/pre/code/ul/ol/li/div/audio/video/img/source, strips all other elements and attributes, allowed classes: mention/hashtag/ invisible/ellipsis/h-*/p-*/u-*/dt-*/e-*), character limits (500 default for composing but accepts and displays longer remote posts with truncation/Read More, hard limit 1MB JSON-LD serialization), thread fetching (up to 5 same-server replies fetched on discovery, inReplyTo recursive fetch, replies not forwarded upstream, 4.4+ fetch-all-replies feature, 4.5 enables by default with AsyncRefresh API), NodeInfo 2.0 at /.well-known/nodeinfo pointing to /nodeinfo/2.0, Mastodon REST API compatibility layer (implemented by Pleroma/Akkoma/GoToSocial/Firefish/ Sharkey/Pixelfed), recent version changes (4.3: attributionDomains/ fediverse:creator, like/boost counts in AP objects, summary as HTML; 4.4: quote post display via FEP-044f, experimental RFC9421 HTTP Message Signatures, FASP support; 4.5: quote post authoring, fetch- all-replies enabled by default, RFC9421 verification standard, AsyncRefresh API), size limits (username 2048, display name 2048 truncated, profile note 20kB, profile fields 50 max, field name/value 2047, poll options 500, emoji shortcode 2048, media descriptions 1500), and common federation gotchas (schema.org incorrect base URI, reply propagation gaps, counter inconsistencies across instances, Block activity sent server-to-server despite spec being client-only, Article type poorly supported, Linked-Data Signatures for public posts disabled in secure mode). Load when the user asks about Mastodon federation; implementing a Mastodon-compatible ActivityPub server; Mastodon's JSON-LD context or toot namespace; how Mastodon handles incoming ActivityPub objects; Mastodon content-type or Accept headers; Mastodon HTTP signatures or authorized fetch; how Mastodon polls work in ActivityPub; how Mastodon edits/Updates federate; how Mastodon boosts/Announces work; Mastodon direct messages in ActivityPub; Mastodon visibility levels; Mastodon HTML sanitization; Mastodon character limits; Mastodon thread fetching; Mastodon NodeInfo; the Mastodon client API compatibility; Mastodon custom emoji federation; Mastodon 4.3/4.4/4.5 changes; or common gotchas when federating with Mastodon.
majiayu000/claude-skill-registry 163
-
dataset-management
Dataset versioning with DVC, HF Datasets processing patterns, data curation workflows, quality checks, and annotation pipeline design with Label Studio and Argilla.
majiayu000/claude-skill-registry 163
-
vue-patterns
Vue 3 Composition API and scoped style policies. Use when writing Vue components, composables, or using :deep() selector. Do NOT use for React/Next.js code -- use react-patterns instead.
majiayu000/claude-skill-registry 163
-
skill-creator
Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.
majiayu000/claude-skill-registry 163
-
Swarm Orchestration
Orchestrate multi-agent swarms with agentic-flow for parallel task execution, dynamic topology, and intelligent coordination. Use when scaling beyond single agents, implementing complex workflows, or building distributed AI systems.
majiayu000/claude-skill-registry 163
-
moai-baas-clerk-ext
Enterprise Clerk Authentication Platform with AI-powered modern identity architecture, Context7 integration, and intelligent user management orchestration for scalable applications
majiayu000/claude-skill-registry 163