cc-sdd
A specification-to-delivery workflow that guides AI agents through design, tasks, implementation, and review.
What it helps with
cc-sdd installs a set of skills for discovery, requirements, design, task breakdown, implementation, and independent review. Claude Code and Codex are marked stable; Cursor, Copilot, Windsurf, OpenCode, Gemini CLI, and Antigravity are marked beta. An agency can use it to keep long-running feature work anchored to approved documents and file boundaries. It can spawn subagents and continue autonomously, so scope, cost, worktree isolation, and review still need human control.
Best fit
Best for developer-led feature delivery with approved specs; not for tiny fixes or teams unable to review autonomous multi-step work.
Setup reality
One-click npx initializer, then needs developer judgment for agent variant, generated skills, boundaries, and autonomous mode.
License context
Permissive licenseBusiness use and modification are generally allowed. Keep required notices and review trademarks, dependencies, and hosted-service terms.MIT. Commercial use and adaptation are permitted with notice retention.
Resource facts
- Use case
- AI Agent Frameworks & Skills
- Industries
- Developers
- Setup
- One-click
- Audience
- Some technical experience
- Capabilities
- No catalog chips
Last verified . Source: github.com