A collection of Claude Code skills, configuration patterns, and research for AI-assisted software engineering. Designed to be copied into ~/.claude/skills/ and adapted to your workflow.
Drop-in workflow skills for Claude Code. Each is a self-contained SKILL.md with triggers, constraints, and structured outputs.
Lifecycle — a complete SDLC pipeline:
| Skill | What it does |
|---|---|
/plan |
Product + engineering review with ADR generation (3 modes: expand, hold, reduce) |
/research |
Structured codebase + external investigation. Returns findings, never implements |
/dev |
Step-by-step implementation with an evaluator subagent reviewing each step's diff |
/execute-plan |
Batch mode: audits plan against code, implements gaps, writes tests |
/test |
Three quality layers: behavioral scoring, property-based testing, mutation testing |
/review |
Pre-landing review for SQL safety, race conditions, trust boundaries |
/ship |
Automated delivery: merge, test, review, version bump, changelog, PR |
/handoff |
Session progress log + generates/updates a docs/ folder |
/debug |
Structured root cause analysis with for-and-against hypothesis testing |
/heal |
Self-healing CI: classifies failures, auto-fixes mechanical issues, escalates the rest |
/drift |
Architectural drift detection across 5 surfaces (ADRs, conventions, deps, docs, schema) |
/retro |
Weekly engineering retrospective with DORA metrics and trend tracking |
/office-hours |
YC-style product coaching before code starts (startup + builder modes) |
/qa |
Browser-based QA: find bugs, fix them, write regression tests |
/qa-map |
Map all live workflows (routes, UI, jobs, IPC) into a structured QA test plan |
Domain — reference implementations for specific stacks:
| Skill | Domain |
|---|---|
/electron-app-deployment |
Build, sign, notarize, and auto-update Electron apps (macOS/Windows/Linux) |
/llm-integration |
LLM provider onboarding: Claude CLI detection, API key management, provider factory |
/stripe-licensing |
Stripe subscriptions + Turso DB + webhook handling for Electron apps |
/user-identification |
Hardware fingerprinting, machine IDs, and tiered licensing |
/vercel-website |
Platform detection, download routing, and rate limiting for landing pages |
/solidity-audit |
Smart contract security audit with 104 vulnerability patterns + Slither/Aderyn |
/solidity-unit-tests |
Foundry + Bulloak BTT methodology with .tree spec files |
Automation — tooling and observability:
| Skill | What it does |
|---|---|
/browse |
Persistent headless Chromium daemon (~100ms/command) for web interaction |
/typefully |
Schedule and manage Twitter/X and LinkedIn posts |
/workflow-discover |
Deprecated — use /qa-map |
/workflow-debug |
Deprecated — use /qa |
Research-backed ideas for extending Claude Code capabilities. Found in claude/extensions/.
Architecture and implementation guide for persistent, personalised AI sessions. Based on primary research across 12+ sources. Found in memory-system/.
Copy the skills you want into your Claude Code skills directory:
# Copy all skills
cp -r claude/skills/* ~/.claude/skills/
# Or copy specific ones
cp -r claude/skills/lifecycle/plan ~/.claude/skills/plan
cp -r claude/skills/lifecycle/dev ~/.claude/skills/devSee claude/GUIDE.md for a decision tree on when to use which skill, with common workflow recipes.
The memory-system/ directory contains research on building persistent AI memory:
- 01-architecture.md — 3-layer memory architecture
- 02-personalisation.md — USER.md design with research backing
- 03-operational-system.md — Decision logging, action items, session orientation
- 04-tooling.md — MCP servers: free vs paid, with config templates
- 05-local-llm-integration.md — Using Ollama/local models as knowledge processors
| Doc | Purpose |
|---|---|
| claude/GUIDE.md | When to use which skill — decision tree + workflow recipes |
| claude/philosophy.md | Configuration philosophy, architecture, key learnings |
| claude/skills/README.md | Full skills reference with cross-skill integration |
| claude/extensions/ | Research proposals for future capabilities |
| memory-system/README.md | Memory system research overview |
MIT
