広告掲載
MCP.so
ログイン
B

Brainstorm MCP Server

@spranab

リポジトリ

基本情報

カテゴリ

AI とエージェント

トランスポート

stdio

作者

spranab

投稿者

Pranab Sarkar

Brainstorm MCP Server について

Multi-model AI brainstorming MCP server. Orchestrates debates between GPT, Gemini, DeepSeek, and Claude with structured synthesis. Includes instant quick mode, multi-model code review with verdicts, and red-team/Socratic styles. Hosted mode needs zero API keys.

Brainstorm MCP Server の設定

以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。

{
  "mcpServers": {
    "brainstorm": {
      "command": "npx",
      "args": [
        "-y",
        "brainstorm-mcp"
      ],
      "env": {
        "OPENAI_API_KEY": "<YOUR_KEY>",
        "GEMINI_API_KEY": "<YOUR_KEY>",
        "DEEPSEEK_API_KEY": "<YOUR_KEY>"
      }
    }
  }
}

Brainstorm MCP Server のツール

7

Run a multi-round brainstorming debate between multiple AI models. IMPORTANT: Before calling this tool, you MUST ask the user to choose a mode: 1. **API mode** — Calls configured providers. These are either HTTP APIs billed per token (OpenAI, Gemini, DeepSeek, ...) or locally installed agent CLIs such as 'claude' and 'codex', which run on the user's existing subscription at no API cost. Call list_providers to see which of each are available. 2. **Hosted mode** — No API keys needed. You execute prompts using sub-agents with models available in your environment (opus/sonnet/haiku, GPT, Gemini, etc.). Same model can be used multiple times — each run produces different perspectives. Present these two options to the user with a one-liner explanation, then proceed based on their choice. For API mode: set mode='api'. When participate=true (default), YOU also participate as a debater alongside external models via brainstorm_respond. For Hosted mode: set mode='hosted'. Ask the user which models to use, then spawn sub-agents for each model, collect responses, and call brainstorm_collect.

Submit YOUR (Claude's) response for the current round of an interactive brainstorm session. After the brainstorm tool returns external models' responses, call this tool with your substantive contribution. Read all responses carefully and engage with specific points — agree, disagree, build upon, or challenge ideas. Do not just summarize. After your response, the next round runs automatically (or synthesis if final round).

Submit collected model responses for a hosted brainstorm session. After receiving prompts from `brainstorm` (mode='hosted') or a previous `brainstorm_collect` call, execute each prompt by spawning a sub-agent for EACH model (use the model parameter to select the right model, e.g., model='sonnet' or model='haiku'). Collect all responses and submit them here. This tool returns either: (1) the next round's prompt to execute, (2) a synthesis prompt for a single model, or (3) the final formatted debate result when complete.

Get instant multi-model perspectives on any question — no debate rounds, no synthesis delay. Fires all models in parallel and returns a compact comparison. Under 10 seconds. Use this for quick second opinions, snap decisions, or when you want diverse perspectives fast. For deeper analysis with multiple rounds and synthesis, use the `brainstorm` tool instead.

Multi-model code review. Pass a diff and get structured findings with severity, file/line references, and a verdict (approve / approve with warnings / needs changes). Multiple models review independently, then findings are synthesized and deduplicated. Use this for PR reviews, code audits, or pre-commit checks.

List all configured AI providers and their default models for brainstorming. Includes locally installed agent CLIs (claude, codex, ...) which run on an existing subscription instead of metered API credits.

Add a new AI provider for brainstorming. Supports any OpenAI-compatible API, or a locally installed agent CLI (kind='cli') that runs on an existing subscription instead of API credits.

概要

What is Brainstorm?

Brainstorm is an MCP server that orchestrates debates between multiple AI models—GPT, Gemini, DeepSeek, Claude—to produce structured synthesis of their perspectives. It is for developers who want more reliable, multi-model insights rather than trusting a single AI.

How to use Brainstorm?

Install with npx -y brainstorm-mcp or npm install -g brainstorm-mcp, then add the server to your .mcp.json (Claude Code) or claude_desktop_config.json (Claude Desktop). Optionally set API keys via environment variables or a config file; hosted mode requires no API keys. Invoke tools such as brainstorm, brainstorm_quick, or brainstorm_review by describing your request to Claude.

Key features of Brainstorm

  • Hosted mode with zero API keys needed
  • API mode with parallel execution across providers
  • Instant multi-model perspectives in under 10 seconds
  • Multi-model code review with severity ratings and verdicts
  • Three debate styles: freeform, red-team, Socratic
  • 3-bullet synthesis verdicts: Recommendation, Key Tradeoffs, Strongest Disagreement
  • Multi-round debates where models critique each other’s responses

Use cases of Brainstorm

  • Compare technologies (e.g., Redis vs PostgreSQL for session storage)
  • Review code diffs for security or correctness issues
  • Decide architectural approaches (e.g., GraphQL vs REST)
  • Brainstorm design ideas with multiple AI perspectives

よくある質問

1

What runtime or dependencies are required?

Node.js and npm. The server is distributed as an npm package and runs via `npx` or global install.

2

Does hosted mode need API keys?

No. Hosted mode uses models available in your environment (Claude Opus/Sonnet/Haiku) via sub-agents; no external API keys are required.

3

Which AI models are supported?

In API mode: OpenAI, Gemini, DeepSeek, Groq, Mistral, Together, and Ollama. Hosted mode supports Claude models (Opus, Sonnet, Haiku).

4

Where does my data go?

In API mode, prompts are sent from your machine to the model providers you configure using your own API keys. Hosted mode makes no external API calls. Sessions are stored in‑memory only with a 10‑minute TTL; no data is written to disk unless you explicitly save results.

5

How does Brainstorm handle errors?

If one model fails (rate limit, timeout), the tool continues with the remaining models and reports which ones failed. Sessions expire after 10 minutes; a clear error message tells you to start a new session.

コメント

「AI とエージェント」の他のコンテンツ

scvd.store

seancrecord

scvd.store is an evidence observatory for agentic commerce. Independent signed observation of what other people's endpoints, artifacts and payments actually did among other things. USDC on Base, Arbitrum One, World Chain

2

AON — Agent Offer Network

AON

AON MCP connects AI agents to real, buyable products and services with live pricing and tracked links—enabling builders to earn developer revenue share from attributed clicks and conversions.

2

Gemina

Gemina Tech Ltd.

Extract, search and tag any document — invoices, receipts, contracts, forms, or your own templates. 13 tools, one sign-in (OAuth 2.1) or API key, free FileTag tier, data residency in the EU, US, Israel or Asia.

hausverbot

eliduarte37

BERGHAIN MCP server — sovereign AI stress-testing mcp / ai / multi-agent / adversarial / stress-test / claude

GetIntel

tarang8811

Public docs + tool manifest for the GetIntel MCP connector (app.getintel.ai)

広告掲載