Skills Catalog
Total Skills: 306 across 32 categories Gateway Skills: 29 (auto-discovered by Claude Code) Agent Skills: 3 (elegant-design, anti-slop, typed-holes-refactor)
Overview
A skill is a directory. {skill}/SKILL.md says what it covers and names the
files beside it; those files are read when they are wanted. That is the whole
structure, and an agent runtime already does the reading, so nothing here
restates the file list a second time.
There used to be a discover-{skill}/ twin for every skill, holding a SKILL.md
that only listed what {skill}/ contained. It was a second address for one
thing, so it is gone.
How It Works
Automatic Discovery
A skill's description is what an agent matches against, so the right ones load
from the words in the request.
Example: "Design a REST API with Postgres" loads api and database; their
reference files are read from there as the work needs them.
Manual Discovery
Browse categories and load skills explicitly:
# List every skill
ls -d skills/*/
# Load a category overview
cat skills/api/INDEX.md
# Load a specific skill
cat skills/api/rest-api-design.md
Gateway Skills by Category
Backend & APIs (7 skills)
Gateway: api
Keywords: API, REST, GraphQL, authentication, authorization, rate limiting
Skills: REST API design, GraphQL schemas, authentication (JWT/OAuth), authorization (RBAC/ABAC), rate limiting, versioning, error handling
cat skills/api/INDEX.md
Databases (8 skills)
Gateway: database
Keywords: database, SQL, PostgreSQL, MongoDB, Redis, query optimization
Skills: PostgreSQL (schema, queries, migrations), MongoDB documents, Redis structures, connection pooling, ORMs, database selection
cat skills/database/INDEX.md
Frontend (8 skills + elegant-design Agent Skill)
Gateway: frontend
Keywords: React, Next.js, UI, components, state management, forms
Skills: React patterns, Next.js App Router, state management, data fetching, forms, accessibility, performance, SEO
Note: elegant-design is a separate Agent Skill for UI design work
cat skills/frontend/INDEX.md
Testing (6 skills)
Gateway: testing
Keywords: testing, tests, unit, integration, e2e, TDD, coverage
Skills: Unit testing, integration testing, e2e testing, TDD, test coverage, performance testing
cat skills/testing/INDEX.md
Diagrams & Visualization (8 skills)
Gateway: diagrams
Keywords: diagram, flowchart, sequence, architecture, ER, UML, Mermaid, visualization
Skills: Flowcharts, sequence diagrams, class/state diagrams, ER diagrams, architecture (C4/block), charts (pie/XY/quadrant/radar), project diagrams (Gantt/timeline), specialized (Git/Sankey/mindmaps)
cat skills/diagrams/INDEX.md
Hanzo Ecosystem (11 skills)
Gateway: hanzo
Keywords: Hanzo, local AI, privacy, Hanzo Node, Hanzo SDK, Hanzo MCP, streaming, ZenLM, training, mining, Zoo Gym, Go, Rust, Python
Skills: Hanzo Node (local inference), Python SDK (multi-provider), Go SDK (concurrent backend), @hanzo/ui (React components), Hanzo MCP (agentic workflows), Hanzo Dev (terminal agent - Rust/JS/Python), Hanzo Live (streaming), ZenLM (local AI models), Zoo Gym (training-free GRPO), Hanzo Desktop (productivity & mining), Hanzo Chat (conversation interface)
cat skills/hanzo/INDEX.md
Why Use Hanzo:
- ⚡ 10-50x faster development: Use SDK abstractions instead of manual API integration
- 🔒 Privacy-first: Local AI via Hanzo Node with selective cloud fallback
- 🎯 Automatic routing: SDK analyzes queries and routes to optimal model
- 💰 Cost optimization: Built-in cost tracking + mine $AI tokens with Desktop
- 🔄 Production-ready: Automatic retries, circuit breaking, caching included
- 🌱 Sustainable: ZenLM models use 95% less energy than cloud AI
- 💚 Wildlife conservation: Zoo Gym by Zoo Labs Foundation (501c3)
Organizations:
- Hanzo AI (github.com/hanzoai): AI tools, SDKs, Desktop, Node, Engine
- ZenLM (github.com/zenlm): ZenLM models on GitHub & HuggingFace
- Zoo AI (github.com/zooai): Zoo Gym training platform
- Zoo Labs Foundation (zoo.ngo): 501(c)(3) for wildlife conservation AI
Machine Learning (33 skills)
Gateway: ml
Keywords: machine learning, ML, AI, models, training, embeddings, transformers
Skills: Model training, inference, embeddings, RAG, evaluation, routing, fine-tuning, PyTorch, TensorFlow, Hugging Face, DSPy with Hanzo SDK, model selection with automatic routing, multi-model orchestration
cat skills/ml/INDEX.md
Enhanced with Hanzo: See dspy-setup.md, llm-model-selection.md, and multi-model-orchestration.md for Hanzo SDK integration patterns.
Mathematics (19 skills)
Gateway: math
Keywords: mathematics, algorithms, linear algebra, calculus, topology, category theory
Skills: Linear algebra, calculus, topology, category theory, set theory, number theory, abstract algebra, graph theory, optimization
cat skills/math/INDEX.md
Elegant Design (Agent Skill)
Gateway: Auto-discovered separately Keywords: design, UI, UX, accessibility, chat, terminal, code display, streaming Skills: World-class accessible interfaces, shadcn/ui, interactive components, design systems
# This is an Agent Skill - loads automatically when doing UI design work
cat skills/elegant-design/SKILL.md
Debugging (14 skills)
Gateway: debugging
Keywords: debugging, GDB, LLDB, profiling, memory leaks, performance
Skills: GDB/LLDB, Python debugging, browser DevTools, remote debugging, memory profiling, concurrency debugging
cat skills/debugging/INDEX.md
Programming Language Theory (13 skills)
Gateway: plt
Keywords: compilers, parsers, type systems, interpreters, LLVM
Skills: Parsing, type systems, code generation, compiler optimizations, LLVM, bytecode, interpreters
cat skills/plt/INDEX.md
Cloud & Serverless (13 skills)
Gateway: cloud
Keywords: cloud, serverless, Modal, AWS, GCP, functions, deployment
Skills: Modal platform, serverless patterns, cloud deployment, scaling, GPU workloads
cat skills/cloud/INDEX.md
Formal Methods (10 skills)
Gateway: formal
Keywords: formal methods, SAT, SMT, Z3, Lean, theorem proving, verification
Skills: SAT solving, SMT theory, Z3 solver, Lean theorem proving, constraint solving, formal verification
cat skills/formal/INDEX.md
Observability (8 skills)
Gateway: observability
Keywords: monitoring, logging, tracing, metrics, distributed tracing, alerts
Skills: Structured logging, metrics instrumentation, distributed tracing, alerting, dashboard design
cat skills/observability/INDEX.md
Build Systems (8 skills)
Gateway: build-systems
Keywords: build systems, Make, CMake, Gradle, Maven, Bazel, compilation
Skills: Make, CMake, Gradle, Maven, Bazel, cross-platform builds, build optimization
cat skills/build-systems/INDEX.md
Caching (7 skills)
Gateway: caching
Keywords: caching, cache, Redis, CDN, HTTP caching, performance
Skills: Caching fundamentals, HTTP caching, CDN edge caching, Redis patterns, cache invalidation, Service Workers, monitoring
cat skills/caching/INDEX.md
Deployment (6 skills)
Gateway: deployment
Keywords: deployment, Netlify, Heroku, CI/CD, production, releases
Skills: Netlify deployment/functions/optimization, Heroku deployment/addons/troubleshooting
cat skills/deployment/INDEX.md
Infrastructure (6 skills)
Gateway: infrastructure
Keywords: infrastructure, Terraform, IaC, Cloudflare Workers, security, DevOps
Skills: Cloudflare Workers, infrastructure security, cost optimization, IaC patterns
cat skills/infrastructure/INDEX.md
Anti-Slop (Agent Skill)
Gateway: Auto-discovered separately Keywords: slop, AI patterns, code quality, content quality Skills: Detect and eliminate generic AI patterns in text, code, and design
# This is an Agent Skill - loads automatically for quality review
cat skills/anti-slop/SKILL.md
Typed Holes Refactor (Agent Skill)
Gateway: Auto-discovered separately Keywords: refactoring, typed holes, incremental refactoring, test-driven Skills: Systematic refactoring using typed holes methodology
# This is an Agent Skill - loads automatically for refactoring work
cat skills/typed-holes-refactor/SKILL.md
Containers (5 skills)
Gateway: containers
Keywords: Docker, containers, Kubernetes, container security, docker-compose
Skills: Dockerfile optimization, docker-compose, container security, networking, registry management
cat skills/containers/INDEX.md
Data Pipelines (5 skills)
Gateway: data
Keywords: ETL, data pipelines, batch processing, stream processing, orchestration
Skills: ETL patterns, batch processing, stream processing, pipeline orchestration, data validation
cat skills/data/INDEX.md
Mobile (4 skills)
Gateway: mobile
Keywords: iOS, Swift, SwiftUI, mobile, SwiftData, concurrency
Skills: SwiftUI architecture, Swift concurrency, iOS networking, iOS testing
cat skills/mobile/INDEX.md
CI/CD (4 skills)
Gateway: cicd
Keywords: CI/CD, GitHub Actions, automation, pipelines
Skills: GitHub Actions, GitLab CI, pipeline patterns, deployment automation
cat skills/cicd/INDEX.md
Realtime (4 skills)
Gateway: realtime
Keywords: realtime, WebSockets, Server-Sent Events, streaming, pub/sub
Skills: WebSocket patterns, SSE, real-time protocols, state synchronization
cat skills/realtime/INDEX.md
WebAssembly (4 skills)
Gateway: wasm
Keywords: WebAssembly, WASM, wasm-pack, Rust to WASM
Skills: WASM fundamentals, Rust to WASM, WASM in browsers, WASI
cat skills/wasm/INDEX.md
eBPF (4 skills)
Gateway: ebpf
Keywords: eBPF, kernel, tracing, networking, BPF
Skills: eBPF fundamentals, tracing, networking, security monitoring
cat skills/ebpf/INDEX.md
Engineering Practices (4 skills)
Gateway: engineering
Keywords: engineering practices, code review, technical leadership
Skills: Code review, documentation, technical leadership, team practices
cat skills/engineering/INDEX.md
Product (4 skills)
Gateway: product
Keywords: product management, roadmap, strategy, prioritization
Skills: Product strategy, roadmap planning, user research, prioritization
cat skills/product/INDEX.md
Collaboration (5 skills)
Gateway: collaboration
Keywords: collaboration, code review, documentation, pair programming
Skills: Code review, documentation, pair programming, team workflows
cat skills/collaboration/INDEX.md
Intermediate Representation (5 skills)
Gateway: ir
Keywords: IR, LLVM IR, intermediate representation, compiler optimizations
Skills: LLVM IR, SSA form, code generation, IR optimizations
cat skills/ir/INDEX.md
Modal (2 skills)
Gateway: modal
Keywords: Modal, serverless functions, cloud deployment
Skills: Modal functions basics, Modal scheduling
cat skills/modal/INDEX.md
Workflow (5 skills)
Gateway: workflow
Keywords: Beads, workflow, task management, synthesis, cleanup, consolidation, roadmap
Skills: Beads workflow, dependency management, multi-session patterns, context strategies, project synthesis
cat skills/workflow/INDEX.md
Quick Reference
Find Skills by Technology
Python: ml, data, testing TypeScript/React: frontend, testing Zig: zig, build-systems Go: cloud, distributed-systems Rust: wasm, formal
Find Skills by Task
Build API: api, database, testing Build Frontend: frontend, elegant-design, api, testing Optimize Performance: caching, database, observability Debug Issues: debugging, observability Deploy Application: deployment, cloud, infrastructure, cicd Data Engineering: data, database, observability
Root-Level Skills
40 skills at root level (not yet categorized):
- Zig skills (zig-*.md)
- TUI skills (bubbletea-.md, ratatui-.md)
- iOS skills (swift*.md, ios-.md, swiftui-.md)
- Modal skills (modal-*.md)
- Networking skills (mosh-.md, mtls-.md, nat-traversal.md, network-resilience-patterns.md, tailscale-vpn.md)
- Discovery/Meta skills (skill-*.md)
- Misc (apache-iceberg.md, duckdb-analytics.md, redpanda-streaming.md)
These will be integrated into appropriate categories in a future update.
Usage Patterns
Automatic (Recommended)
Just start working. Claude Code will auto-activate relevant gateway skills.
Browse Then Load
- Review this README for category overview
- Load category index:
cat skills/{category}/INDEX.md - Load specific skill:
cat skills/{category}/{skill}.md
Search by Keyword
# Find skills matching a keyword
grep -r "keyword" skills/*/INDEX.md
# Find a skill by what it says it covers
grep "description:" skills/*/SKILL.md
Architecture Benefits
✅ Context Efficient: 60-84% reduction in tokens loaded ✅ Auto-Discovery: Keywords trigger relevant gateway skills ✅ Progressive Loading: Load only what you need, when you need it ✅ Scalable: Can grow to 1000+ skills with same footprint ✅ Maintainable: Small, focused files instead of monolithic index ✅ Flexible: Works for automatic AND manual discovery
See also:
skill-repo-discovery.md- Discover skills for a new repositoryskill-prompt-discovery.md- Discover skills from user promptsCLAUDE.md- Complete development guidelines with Work Plan Protocol