ccg
starsforks
How to install ccg?
jmstar85/oh-my-githubcopilot/ccg
$npx -y skills add jmstar85/oh-my-githubcopilot --skill ccgClaude-Codex-Gemini tri-model orchestration for multi-perspective analysis. Activate when user says: ccg, tri-model, three models, multi-model, cross-validate, get multiple opinions, compare models.
SKILL.md
| 1 | # CCG - Claude-Codex-Gemini Tri-Model Orchestration |
| 2 | |
| 3 | Route a task through three AI models in parallel, then synthesize their outputs into one unified answer. |
| 4 | |
| 5 | ## When to Use |
| 6 | - Backend/analysis + frontend/UI work in one request |
| 7 | - Code review from multiple perspectives |
| 8 | - Cross-validation where models may disagree |
| 9 | - Fast parallel input without full team orchestration |
| 10 | |
| 11 | ## When NOT to Use |
| 12 | - Simple, straightforward tasks → execute directly |
| 13 | - Already clear on approach → use `/omg-autopilot` |
| 14 | - Need coordinated multi-agent work → use `/team` |
| 15 | |
| 16 | ## Requirements |
| 17 | |
| 18 | - **Codex CLI**: `npm install -g @openai/codex` |
| 19 | - **Gemini CLI**: `npm install -g @google/gemini-cli` |
| 20 | - If either CLI is unavailable, continue with whichever provider works |
| 21 | |
| 22 | ## Execution Protocol |
| 23 | |
| 24 | ### 1. Decompose Request |
| 25 | |
| 26 | Split the user request into: |
| 27 | - **Codex prompt**: architecture, correctness, backend, risks, test strategy |
| 28 | - **Gemini prompt**: UX/content clarity, alternatives, edge-case usability, docs polish |
| 29 | - **Synthesis plan**: how to reconcile conflicts |
| 30 | |
| 31 | ### 2. Invoke Advisors |
| 32 | |
| 33 | Run both advisors via CLI in parallel: |
| 34 | |
| 35 | ```bash |
| 36 | # Run in terminal |
| 37 | codex "<codex prompt>" |
| 38 | gemini "<gemini prompt>" |
| 39 | ``` |
| 40 | |
| 41 | Or via VS Code's `selectChatModels()` API if available: |
| 42 | ``` |
| 43 | Promise.all([ |
| 44 | model_openai.sendRequest(codex_prompt), |
| 45 | model_google.sendRequest(gemini_prompt) |
| 46 | ]) |
| 47 | ``` |
| 48 | |
| 49 | ### 3. Collect Results |
| 50 | |
| 51 | Gather outputs from both advisors. |
| 52 | |
| 53 | ### 4. Synthesize |
| 54 | |
| 55 | Return one unified answer with: |
| 56 | - **Agreed** recommendations |
| 57 | - **Conflicting** recommendations (explicitly called out) |
| 58 | - **Chosen** final direction + rationale |
| 59 | - **Action** checklist |
| 60 | |
| 61 | ## Fallbacks |
| 62 | |
| 63 | | Scenario | Action | |
| 64 | |----------|--------| |
| 65 | | One provider unavailable | Continue with available + Claude synthesis | |
| 66 | | Both unavailable | Fall back to Claude-only answer | |
| 67 | |
| 68 | ## Example |
| 69 | |
| 70 | ``` |
| 71 | /ccg Review this PR - architecture/security via Codex and UX/readability via Gemini |
| 72 | ``` |
| 73 | |
| 74 | Output: |
| 75 | ``` |
| 76 | === CCG Synthesis === |
| 77 | |
| 78 | ## Agreed |
| 79 | - Authentication middleware needs rate limiting |
| 80 | - Error messages should be more user-friendly |
| 81 | |
| 82 | ## Conflicting |
| 83 | - Codex: Use middleware pattern for validation |
| 84 | - Gemini: Use inline validation for simplicity |
| 85 | → Chosen: Middleware pattern (consistency with existing codebase) |
| 86 | |
| 87 | ## Action Checklist |
| 88 | - [ ] Add rate limiting middleware |
| 89 | - [ ] Improve error messages in auth flow |
| 90 | - [ ] Extract validation to middleware layer |
| 91 | ``` |
People who liked this also viewed
6 picksType
- find-skillsHelps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skSkillsJul 2026
- grill-meA relentless interview to sharpen a plan or design.SkillsJul 2026
- vercel-react-best-practicesVercel's official collection of agent skillsSkillsJul 2026
- improve-codebase-architectureScan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.SkillsJul 2026
- setup-matt-pocock-skillsConfigure this repo for the engineering skills — set up its issue tracker, triage label vocabulary, and domain doc layout. Run once before first use of the other engineering skills.SkillsJul 2026
- triageMove issues and external PRs through a state machine of triage roles — categorise, verify, grill if needed, and write agent-ready briefs.SkillsJul 2026
