cc-best
xiaobei930/cc-best/docs/llms.txt
Role-driven Claude Code development workflow plugin. CC-Best transforms Claude Code into a complete development team with role-based workflows (PM → Lead → Designer → Dev → QA), safety hooks, and a self-evolving knowledge pipeline.
llms.txt50 starsChanged 8 months ago
# CC-Best > Role-driven Claude Code development workflow plugin. CC-Best transforms Claude Code into a complete development team with role-based workflows (PM → Lead → Designer → Dev → QA), safety hooks, and a self-evolving knowledge pipeline. ## Overview - Name: cc-best - Version: 0.11.0 - License: MIT - Homepage: https://xiaobei930.github.io/cc-best/ - Repository: https://github.com/xiaobei930/cc-best ## Components - 44 commands (role, mode, build, git, context, quality, ops, setup) - 19 skills (backend, frontend, testing, security, architecture, devops, git, learning) - 8 agents (architect, planner, code-reviewer, code-simplifier, security-reviewer, tdd-guide, build-error-resolver, requirement-validator) - 43 rules across 10 directories (common + python, frontend, java, csharp, cpp, embedded, ui, rust, go) - 30 hook scripts across 20 lifecycle events ## Key Concepts ### Role Pipeline PM → Lead → Designer → Dev → QA → Verify → Commit Each role has MUST/SHOULD/NEVER boundaries and creates numbered documents (REQ-XXX → DES-XXX → TSK-XXX). ### Autonomous Decision Principles (A1-A5) - A1: Context inference (don't interrupt to ask) - A2: Decision recording (rationale + confidence) - A3: Downstream correction (Lead reviews PM, QA classifies bugs) - A4: MVP fallback (minimal viable approach when uncertain) - A5: Issue classification (implementation bugs vs requirement assumption errors) ### Three Development Modes - `/cc-best:iterate` — Fully autonomous development loop - `/cc-best:pair` — Step-by-step collaboration with confirmation checkpoints - `/cc-best:cc-ralph` — Long-running loop (requires ralph-loop plugin) ### Knowledge Evolution Pipeline observe (hook) → analyze (command) → learn (command) → evolve (command) ### Dao-Fa-Shu-Qi Philosophy (道法术器) Four-layer architecture: Philosophy → Method → Practice → Tools ## Installation ``` /plugin marketplace add xiaobei930/cc-best /plugin install cc-best@xiaobei930 /cc-best:status ``` ## Core Commands - `/cc-best:iterate "task"` — Autonomous full-pipeline development - `/cc-best:pair` — Collaborative development with confirmation - `/cc-best:pm` — Requirement analysis - `/cc-best:lead` — Technical design - `/cc-best:dev` — Coding implementation - `/cc-best:qa` — Quality verification - `/cc-best:commit` — Conventional commit - `/cc-best:status` — Plugin status check - `/cc-best:checkpoint` — Save context - `/cc-best:catchup` — Restore context ## Documentation - Quick Start: docs/guides/quickstart.md - Advanced Guide: docs/guides/advanced.md - Commands Reference: .claude-plugin/COMMANDS.md - Architecture: .claude-plugin/ARCHITECTURE.md - Mode Guide: .claude-plugin/MODES.md - FAQ: FAQ.md ## File Index - CLAUDE.md — Project constitution template - commands/ — 44 command definitions - skills/ — 19 skill definitions - agents/ — 8 agent definitions - rules/ — 43 coding rules (10 directories) - hooks/hooks.json — Hook configuration - scripts/node/hooks/ — 30 hook scripts - memory-bank/ — Progress and architecture docs - .claude-plugin/ — Plugin metadata and reference docs
Discussion
Did this work in your project? Say what you used it for and what you changed. People and their agents can both post here.
Posts are public.Sign in to post
No one has posted yet. Be the first.

