Agent skills

Skills you can use with AI coding agents, indexed from public GitHub repositories.

Latest Most Popular

test-before-commit Enforces running unit tests and linting before committing code changes in the unsealed-networks repository
majiayu000/claude-skill-registry 163
psr12-moodle Automatically validate and fix Moodle PHP code for PSR-12 compliance with Moodle-specific exceptions (lowercase_with_underscores naming, frankenstyle prefixes). Activates when working with Moodle plugin PHP files or when code standards issues are detected.
majiayu000/claude-skill-registry 163
mlx-dev Write correct, idiomatic Apple MLX code for Apple Silicon ML. Use when working with MLX arrays, neural networks, training loops, lazy evaluation, unified memory, mx.eval, mx.compile, Metal GPU, memory optimization, quantization, or Apple Silicon performance. Covers critical API differences from PyTorch/NumPy, array indexing gotchas (lists must be mx.array, slices create copies), NHWC format for Conv2d, __call__ not forward(), float64 CPU-only, mlx-lm integration, and debugging patterns.
majiayu000/claude-skill-registry 163
agent-ready-eval Evaluate a codebase for agent-friendliness based on autonomous agent best practices. Use when asked to "evaluate for agents", "check agent readiness", "audit for autonomous execution", "assess agent-friendliness", or when reviewing infrastructure for unattended agent operation. Also use when asked about making a codebase more suitable for AI agents or autonomous workflows.
majiayu000/claude-skill-registry 163
executing-ai-development-workflow Execute a comprehensive AI-driven development workflow with planning, implementation, multi-layer review (Sub-agents + /review + CodeRabbit CLI), automated fixes, and PR creation. Use when implementing new features, performing large refactorings, developing security-critical features, or when the user mentions "AI開発ワークフロー", "AI development workflow", or "計画的に実装".
majiayu000/claude-skill-registry 163
repo-explore Clone and explore external GitHub repositories to understand how libraries, frameworks, or dependencies work. Use when user provides a GitHub URL (github.com/owner/repo), asks "how does X library work", wants to look at source code for a dependency, asks about implementation details of an external package, or says "explore", "look at", or "check out" a repository. Automatically checks out the matching version tag when the repo is a dependency in the current project.
majiayu000/claude-skill-registry 163
threedee-parametric-tazomatalax-threedee
majiayu000/claude-skill-registry 163
ptk-completion This skill should be used when the user asks about "prompt_toolkit completion", "Completer", "WordCompleter", "FuzzyCompleter", "NestedCompleter", "PathCompleter", "auto-complete", "tab completion", "completion menu", "CompletionsMenu", or needs to implement auto-completion in prompt_toolkit applications.
majiayu000/claude-skill-registry 163
scanning-for-secrets Detect exposed secrets, API keys, and credentials in code. Use when auditing for secret leaks. Trigger with 'scan for secrets', 'find exposed keys', or 'check credentials'.
majiayu000/claude-skill-registry 163
generate-tags Generate tags for a blog post.
majiayu000/claude-skill-registry 163
unknown-majiayu000-claude-skill-registr-76
majiayu000/claude-skill-registry 163
python-conventions Work on Python experiments in packages/python_viterbo. Use for layout conventions, stage entrypoints, lint/test commands, and asset/plot handling.
majiayu000/claude-skill-registry 163
webapp-wow-mechanic-creator Design innovative game mechanics using web browser APIs that create "WOW moments" impossible in other platforms. Implements 5 surprise types (rule-breaking, meta-awareness, recursive-explosion, hidden-layer, system-hacking) using LocalStorage, Canvas, Notifications, DevTools, and browser tab manipulation. Use when creating surprising game mechanics, implementing meta-game elements, or designing memorable interactive moments for web games.
majiayu000/claude-skill-registry 163
skills-pocketllm-pocketllm
majiayu000/claude-skill-registry 163
cpp-interface-purpose-comments Add a standard Purpose comment above each pure virtual method declaration in C++ interface header files (.h/.hpp). Use when asked to batch annotate interface classes or enforce a consistent comment block above `virtual ... = 0;` declarations.
majiayu000/claude-skill-registry 163
cache-cost-tracking LLM cost tracking with Langfuse for cached responses. Use when monitoring cache effectiveness, tracking cost savings, or attributing costs to agents in multi-agent systems.
majiayu000/claude-skill-registry 163
Architecture Analyst Expert in software architecture analysis, pattern identification, and architectural quality assessment
majiayu000/claude-skill-registry 163
codex-taskloop-plugin Run in-session task loops via the codex-taskloop-plugin MCP server and stop hook.
majiayu000/claude-skill-registry 163
consul-service-discovery Consul Service Discovery - Auto-activating skill for DevOps Advanced. Triggers on: consul service discovery, consul service discovery Part of the DevOps Advanced skill category.
majiayu000/claude-skill-registry 163
conventions Load coding conventions and style guidelines when writing or reviewing code. Ensures consistency with project standards.
majiayu000/claude-skill-registry 163
cloudflare-vectorize Build semantic search with Cloudflare Vectorize V2 (Sept 2024 GA). Covers V2 breaking changes: async mutations, 5M vectors/index (was 200K), 31ms latency (was 549ms), returnMetadata enum, and V1 deprecation (Dec 2024). Use when: migrating V1→V2, handling async mutations with mutationId, creating metadata indexes before insert, or troubleshooting "returnMetadata must be 'all'", V2 timing issues, metadata index errors, dimension mismatches.
majiayu000/claude-skill-registry 163
implement-orpc-api Implement a new handler and attach to the oRPC router
majiayu000/claude-skill-registry 163
schedule-verification Human verification checklist for generated schedules. Use when reviewing Block 10 or any generated schedule to ensure it makes operational sense. Covers FMIT, call, Night Float, clinic days, and absence handling.
majiayu000/claude-skill-registry 163
gastown Multi-agent orchestrator for Claude Code. Use when user mentions gastown, gas town, gt commands, bd commands, convoys, polecats, crew, rigs, slinging work, multi-agent coordination, beads, hooks, molecules, workflows, the witness, the mayor, the refinery, the deacon, dogs, escalation, or wants to run multiple AI agents on projects simultaneously. Handles installation, workspace setup, work tracking, agent lifecycle, crash recovery, and all gt/bd CLI operations.
majiayu000/claude-skill-registry 163