Deliberation
Deliberation MCP is a MCP server that delivers second opinions from GPT, Gemini, Grok, and 400+ OpenRouter models inside Claude Code and other MCP hosts.
About
Deliberation MCP exposes a stdio Model Context Protocol server that federates second opinions from major frontier models and the broader OpenRouter catalog into the same workspace where you already code. Developers often standardize on Claude Code or Cursor for implementation but still want Gemini, GPT, or Grok to stress-test a design doc, threat model, or patch series before shipping. This server packages that multi-model loop as MCP tools rather than copy-pasting prompts across browser tabs. Typical flow: finish a plan or diff in your main agent, invoke deliberation with a focused question, and compare answers before you merge. Complexity is intermediate because you must configure OpenRouter credentials and choose models intentionally. It is deliberation infrastructure, not autonomous code review on every file save—use it at decision points where a wrong call is expensive.
- stdio npm package @antonbabenko/deliberation-mcp at catalog version 3.5.3
- Routes deliberation requests through OpenRouter to GPT, Gemini, Grok, and 400+ other models
- Works inside Claude Code and any MCP-capable host as a second-opinion sidecar
- Useful before merging risky changes, refactors, or infra edits you only validated with one model
- Keeps your primary session in one agent while alternate models challenge assumptions
Deliberation by the numbers
- Data as of Jul 7, 2026 (Skillselion catalog sync)
claude mcp add deliberation -- npx -y @antonbabenko/deliberation-mcpAdd your badge
Show developers this MCP server is listed on Skillselion. Paste this into your README.
| Package | @antonbabenko/deliberation-mcp |
|---|---|
| Transport | STDIO |
| Auth | None |
What it does
Pull structured second opinions from GPT, Gemini, Grok, and hundreds of OpenRouter models while you stay in Claude Code or another MCP host.
Who is it for?
Best when you're in Claude Code and want fast multi-model critique on plans, diffs, and runbooks without leaving MCP.
Skip if: Skip if you cannot use OpenRouter, need guaranteed offline review, or want a fully automated CI gate with no API cost.
What you get
After you wire Deliberation MCP, you can query alternate models from the same session and incorporate dissenting views before you ship.
- Cross-model answers to focused technical questions from the active IDE session
- Documented dissent or agreement you can paste into PR description or plan
- Repeatable deliberation workflow without switching chat products
By the numbers
- Package @antonbabenko/deliberation-mcp version 3.5.3
- Transport: stdio via npm registry
- OpenRouter access to 400+ models including GPT, Gemini, and Grok families
Recommended MCP Servers
0Latency MemoryPersistent memory layer for AI agents.
0nMCP — Universal AI API OrchestratorUniversal AI API Orchestrator — 1,554 tools, 96 services.
0xHumans Protocol MCPMCP for AI agents: financing, skills, lending on Base1k Patient Mcp1k patient MCP server
1trippulse1trip PULSE: 21-tool AI travel planner.
3D AI Agent Avatar3D AI Agent Avatar — render any GLB, give it a Solana wallet, a voice, and pump.fun powers.89How it compares
MCP multi-model deliberation layer, not a single-vendor chat skill or static linter.
FAQ
Who is Deliberation MCP for?
Developers using Claude Code, Cursor, or generic MCP hosts who want GPT, Gemini, Grok, or other OpenRouter models as a second opinion on technical decisions.
When should I use Deliberation MCP?
Use it during Ship review when a change is hard to unwind—security-sensitive edits, architecture shifts, or ambiguous requirements—before merge or deploy.
How do I add Deliberation MCP to my agent?
Install @antonbabenko/deliberation-mcp from npm, register the stdio server in your MCP host config, and provide OpenRouter API access per the package README.