広告掲載
MCP.so
ログイン

Heuresis MCP Server

@ToremLabs

リポジトリ

基本情報

カテゴリ

その他

ライセンス

AGPL-3.0

トランスポート

stdio

作者

ToremLabs

投稿者

Torem Labs

Heuresis MCP Server について

概要はまだありません

#ideation#knowledge-graph#brainstorming#triz#productivity#research

Heuresis 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": {
    "heuresis": {
      "command": "npx",
      "args": [
        "-y",
        "@heuresis/mcp@latest"
      ]
    }
  }
}

Heuresis MCP Server のツール

ツール一覧はまだありません

この掲載には表示できるツール一覧がまだありません。サーバーにツールがないという意味ではありません。対応機能はプロジェクトのドキュメントをご確認ください。

上のボタンで一時的なサンドボックスでこのサーバーを実行し、ツール一覧を取得できます。

概要

What is Heuresis?

Heuresis is a Model Context Protocol (MCP) server that exposes a Heuresis workspace to any MCP‑capable client—Claude Desktop, Claude Code, Cursor, Windsurf, or custom agents. It logs into the user’s Heuresis account, communicates with the same Supabase project the webapp uses, and respects the same Row‑Level Security (RLS). The webapp and the MCP server are two front‑ends to one cloud workspace.

How to use Heuresis?

Install the package globally (npm install -g @heuresis/mcp) or run it on‑demand with npx -y @heuresis/mcp. First link the machine to your Heuresis account using npx -y -p @heuresis/mcp heuresis-mcp login, then configure your MCP client (e.g., Claude Desktop or a .mcp.json file) to run the @heuresis/mcp server. For ephemeral environments (CI, cloud agents), set environment variables HEURESIS_EMAIL, HEURESIS_PASSWORD, and HEURESIS_ANON_KEY to sign in fresh on every boot.

Key features of Heuresis

  • 34 tools: 31 data tools and 3 operator tools
  • Live sync via Supabase Realtime (no manual refresh)
  • Device‑code login for personal machines
  • Email/password authentication for headless/CI usage
  • Respects workspace Row‑Level Security (RLS)
  • All writes stamped with origin='mcp' for session audit

Use cases of Heuresis

  • Ideate and manage workspace content from Claude Desktop or similar clients
  • Automate concept editing, linking, and project creation in CI pipelines
  • Collaborate with the webapp — changes sync instantly via Realtime
  • Search concepts and explore project graphs through natural‑language agent prompts
  • Run creative operators (Branch, Matrix, ASIT, TRIZ, etc.) from an agent

よくある質問

1

How do I authenticate the MCP server?

On a personal machine, run `heuresis-mcp login` and follow the device‑code URL to confirm in your browser. In ephemeral environments (CI, cloud agents), set `HEURESIS_EMAIL`, `HEURESIS_PASSWORD`, and `HEURESIS_ANON_KEY` environment variables to sign in fresh on each boot.

2

Can I unlink a machine?

Yes — run `heuresis-mcp logout` to delete the credentials file locally, or go to Settings ▸ Connected devices in the Heuresis webapp to revoke a device remotely.

3

What tools are available?

The server exposes 34 tools: 10 read tools (e.g., `get_workspace_summary`, `list_projects`, `search_concepts`), 21 write tools (e.g., `add_concept`, `link_concepts`, `create_idea`), and 3 operator‑run tools (`run_operator`, `run_operator_and_commit`, `expand_concept`).

4

Does the MCP server sync with the webapp?

Yes — by default the server subscribes to Supabase Realtime and notifies the client whenever rows change in `nodes`, `edges`, `projects`, or `ideas`. Edits made in the webapp appear without a manual refresh.

5

Is there a legacy read‑only mode?

Yes, but it is deprecated. If no credentials file exists and the `HEURESIS_SNAPSHOT` env var points to a JSON export, the server reads that file and exposes a reduced set of read‑only tools. This mode will be removed in a later release.

コメント

「その他」の他のコンテンツ

Hindsight

vectorize-io

Hindsight: Agent Memory That Learns

50.4K

🚀 Model Context Protocol (MCP) Curriculum for Beginners

microsoft

This open-source curriculum introduces the fundamentals of Model Context Protocol (MCP) through real-world, cross-language examples in .NET, Java, TypeScript, JavaScript, Rust and Python. Designed for developers, it focuses on practical techniques for building modular, scalable,

16.7K
P

PickFu MCP

PickFu MCP is a Model Context Protocol (MCP) server that connects AI assistants, developers, marketers, product managers, and AI agents to PickFu's consumer research platform. It enables users to create surveys, conduct

EverArt

modelcontextprotocol

Model Context Protocol Servers

88.1K

Klarix Intelligence Engine

Klarix

B2B competitive intelligence, live sales battlecards, 5-axis ICP prospect fit scoring, and tech stack teardowns over MCP and REST with cited public-web evidence.

広告掲載