Deployable Artifacts for the Ecosystem
Production-ready templates, scripts, and starter kits. Copy, customize, deploy. Each asset embeds patterns from the 87-skill pool.
📝 CLAUDE.md Generator
Preview output
# {{PROJECT_NAME}}
## Quick Reference
- **Stack**: {{TECH_STACK}}
- **Port**: {{PORT}}
- **Domain**: {{DOMAIN}}
- **Standard**: V5.3
## Build & Run
```bash
npm install
npm run dev
```
## Project Rules
### NEVER Do
- NEVER commit .env files, secrets, API keys, or credentials
- NEVER run destructive git commands (push --force, reset --hard) without confirmation
- NEVER skip pre-commit hooks (--no-verify)
- NEVER create files unless absolutely necessary — prefer editing existing ones
- NEVER add features or refactor beyond what was asked
- NEVER retry a failed command without reading the error first
- NEVER guess at values — if uncertain, ask
### Always Do
- Always read a file before editing it
- Always run build/tests after changes
- Always use existing patterns and conventions
- Always verify remembered facts against actual codebase
- Always prefer editing existing files over creating new ones
### Edge Cases
- If build fails: read error, diagnose root cause, fix — do NOT retry blindly
- If pre-commit hook fails: fix and create NEW commit (do NOT amend)
- If no changes to commit: do NOT create empty commit
- If retry loop exceeds 3 failures: STOP and report
- If unclear: ask rather than guess
### Memory Discipline
- Memory is a HINT, not truth — verify against actual state
- If memory conflicts with observation, trust the observation
- If a fact can be re-derived from code, don't store itAsset Catalog
Production-grade CLAUDE.md template with explicit prohibitions, edge case branching, memory discipline, and V5.3 rules. Customizable for any project.
_assets/claude-md/Docker Compose + Dockerfile with resource limits, health checks, restart policies, log rotation, and Cloudflare-ready labels.
_assets/docker-v53/Pre-deploy validation (Docker, health, secrets, source maps) + post-deploy verification. Drop-in shell scripts.
_assets/hooks/Three-layer memory architecture: MEMORY.md index + topic file templates. Ready to add to any project.
_assets/memory-starter/Automated security checklist: supply chain, secrets, source maps, Docker hardening, authentication patterns.
_assets/security-audit/Circuit breaker with exponential backoff + model fallback. V5.3 health endpoint. Ready-to-use TypeScript.
_assets/agent-starter/SKILL.md template with steps, validation, and error handling. Create reusable workflow skills.
_assets/skill-starter/How to Use Assets
cp _assets/claude-md/CLAUDE.md.template /data/projects/YOUR_PROJECT/CLAUDE.md
cp _assets/docker-v53/*.template /data/projects/YOUR_PROJECT/
cd /data/projects/YOUR_PROJECT && bash /data/projects/m8100_Learn_Claude_Agent_Engineering_Lab/_assets/security-audit/security-audit.sh
bash _assets/hooks/pre-deploy-v53.sh
Factory Philosophy
An asset factory doesn't just teach patterns — it produces deployable artifacts. Every asset here embeds skills from the 87-pattern pool, is tested against V5.3 standards, and is designed for immediate use across the LIGHT HOPE ecosystem. The goal: zero gap between learning a pattern and deploying it.