#9393rd of 149 in Skills, plugins and rules
ccg-workflow
Multi-model workflow layer for Claude Code, delegating to Codex, Grok and Kimi
Documentation ↗
(opens in a new tab)Website ↗
(opens in a new tab)Repository on GitHub ↗
(opens in a new tab)
- Stars
- 5.9k
- License
- MIT
- Last commit
- Oct 2026
- Language
- TypeScript
4,228 on 25 Mar 20265,945 on 14 Oct 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
CCG installs slash commands, hooks, skills and a compiled Go bridge (codeagent-wrapper) into Claude Code. Claude stays the orchestrator and picks one of 10 strategies per task, sending analysis and review to external models such as Codex, Grok, Kimi Code and Antigravity. Hooks inject task state every turn so context survives compaction, and tasks persist under .ccg/tasks.
Who it is for: Claude Code users who want to route work to other models
Strengths
- 10 built-in strategies, from direct-fix to full-collaborate with dual-model review
- Hooks re-inject task state every turn and after context compaction
- Skills install individually; doctor --json outputs a versioned health report
- Native plugin option installs skills only, with no binary
Weaknesses
- Multi-model commands ship only via the npx install, not the native plugin
- Requires Claude Code CLI; external models need their own CLIs installed
- Installs a compiled Go binary into ~/.claude/bin as the model bridge
- README mixes in sponsor blocks and unrelated projects
What it needs
- no GPU
- Needs Node.js 20+, Claude Code CLI, Codex CLI (optional), Grok CLI (optional), Kimi Code CLI (optional), Antigravity (optional)
- Models: Claude, Codex (OpenAI), Grok (xAI), Kimi Code (Moonshot), Antigravity
Also in Skills, plugins and rules
See all 149| Rank | Project | Score |
|---|---|---|
| 1 | cavemanCuts token use in coding agents with terse replies and compressed inputs | 95 out of 100 |
| 2 | skillsAgent skills for planning, TDD, debugging and code review | 94 out of 100 |
| 3 | eccAgents, skills, hooks and rules bundle for Claude Code, Codex and other harnesses | 94 out of 100 |
| #4 | impeccableDesign skill with 24 commands and a detector for AI-generated frontends | 94 out of 100 |
| #5 | ponytailSkill prompt that makes coding agents write less code and add tests | 92 out of 100 |