Stats
Actions
Tags
From dev-plugin
Use this skill when you need to delegate work to other AI models via OpenCode, get feedback from multiple models, or orchestrate collaborative multi-agent workflows. Claude Code stays in control and decides what to ask, which models to use, and how to use their responses.
How this skill is triggered — by the user, by Claude, or both
Slash command
/dev-plugin:ask-opencodeThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
A simple bridge between Claude Code and other AI models via OpenCode CLI.
A simple bridge between Claude Code and other AI models via OpenCode CLI.
Lets Claude Code communicate with other AI models (Gemini, Codex, GPT-5, etc.) through OpenCode.
You (Claude Code) control:
Install OpenCode:
curl -fsSL https://raw.githubusercontent.com/opencode-ai/opencode/refs/heads/main/install | bash
Configure:
opencode /connect # Add your API keys
Test:
opencode run "Hello"
opencode run -m "github-copilot/gpt-5.1-codex" "Your prompt here"
# Ask Gemini
opencode run -m "github-copilot/gemini-2.5-pro" "Your prompt"
# Ask Codex
opencode run -m "github-copilot/gpt-5.1-codex" "Your prompt"
# Ask Claude
opencode run -m "github-copilot/claude-sonnet-4.5" "Your prompt"
opencode models
Note: Requires GitHub Copilot access with OpenCode integration.
You: "I need tests for this function"
→ opencode run -m "github-copilot/gpt-5.1-codex" "Write tests for: [code]"
← Review response
You: Use the tests
You: "I wrote this code"
→ opencode run -m "github-copilot/claude-opus-4.6" "Review: [code]"
← Review response
You: Fix issues
You: "Which approach is better?"
→ Ask model 1
→ Ask model 2
→ Ask model 3
← Read all responses
You: Decide based on consensus
You: "Need algorithm + API"
→ opencode run -m "github-copilot/gemini-2.5-pro" "Design algorithm"
→ opencode run -m "github-copilot/gpt-5.1-codex" "Build API"
← Review both
You: Combine them
You: "Initial solution"
→ Ask for feedback
← Get feedback
You: Improve solution
→ Ask again
← Get validation
You: Done
Available GitHub Copilot Models:
| Model | Best For |
|---|---|
github-copilot/gpt-5.1-codex | Code generation, refactoring |
github-copilot/gpt-5.1-codex-max | Complex code tasks |
github-copilot/gpt-5.2-codex | Advanced code analysis |
github-copilot/claude-sonnet-4.5 | Balanced reasoning & code |
github-copilot/claude-opus-4.6 | Deep analysis, security review |
github-copilot/gemini-2.5-pro | Complex reasoning |
github-copilot/gpt-5.2 | General purpose |
github-copilot/gpt-5.1 | Fast responses |
Run opencode models to see all available models.
Collaborative Spec Writing:
opencode run -m "github-copilot/claude-sonnet-4.5" "Review this spec: [spec]"opencode run -m "github-copilot/gpt-5.1-codex" "Is this implementable: [spec]"That's it. Simple orchestration, you control everything.
npx claudepluginhub rsolipuram/claude-code-plugins --plugin dev-pluginInternal guidance for composing Copilot prompts for coding, review, diagnosis, and research tasks across GPT-5.4, GPT-5.3-Codex, and Gemini 3.1 Pro inside the Copilot Claude Code plugin
Delegates tasks to GitHub Copilot CLI non-interactively with multi-model support (Claude, GPT, Gemini), permission controls, output sharing, and session resume. Useful for Copilot delegation requests, model comparisons, or programmatic runs.
Invokes OpenAI Codex and Google Gemini CLIs via Bash for second opinions, code reviews, and alternative analysis. Useful when users request external AI verification or explicitly say 'ask codex' or 'ask gemini'.