Reusable Claude Code infrastructure with agent coordination, design system, and OpenSpec integration. Project-agnostic and works with any language/framework.
.claude/
├── agents/ # 15 specialized agent definitions
├── commands/ # Slash commands for common workflows
├── skills/ # Reusable skill packages
└── reports/ # Generated reports and registries
openspec/ # Spec-driven development
| Command | Purpose |
|---|---|
/agents:review [path] |
Code review |
/agents:security [path] |
Security scan |
/agents:coverage |
Test coverage analysis |
/agents:ci |
CI pipeline (lint, type-check, test) |
/test |
Run tests (auto-detects runner) |
/release |
Version bump and release |
/rfc |
Create/review design documents |
/review-full [path] |
Multi-level review |
Commands auto-detect project type from config files:
| Config File | Project Type |
|---|---|
pyproject.toml |
Python (uses UV) |
package.json |
Node.js |
Cargo.toml |
Rust |
go.mod |
Go |
- UV - For running infrastructure scripts
- Claude Code CLI - For agent execution
Copy the .claude/ and openspec/ directories into your project. The infrastructure will adapt to your project's toolchain automatically.
See CLAUDE.md for AI assistant instructions.