MCP.so
ログイン

humanmcp

@Syuparn

humanmcp について

You are the MCP server!

基本情報

カテゴリ

その他

ライセンス

MIT

ランタイム

go

トランスポート

stdio

公開者

Syuparn

設定

標準の設定はありません

このサーバーの README には解析可能な MCP 設定ブロックが含まれていません。インストール手順はリポジトリをご確認ください。

リポジトリ

ツール

ツールは検出されませんでした

ツールは README から自動的に抽出されます。メンテナーは ## Tools という見出しの下に記載することで、このタブに反映できます。

概要

What is humanmcp?

humanmcp is a joke MCP server that requires a human to manually process all requests and responses. Instead of automating tools, the user reads incoming JSON-RPC requests from in.txt and writes the corresponding responses to out.txt. It is intended as a humorous parody of automated MCP servers.

How to use humanmcp?

  1. Download the server binary from the Releases page.
  2. Add the server to your MCP host settings (e.g., VSCode) with the command pointing to the binary.
  3. Monitor in.txt for incoming requests using tail -f in.txt.
  4. Manually craft and write the correct JSON-RPC response to out.txt using echo '...' > out.txt or similar.

Key features of humanmcp

  • 100% manual operation – no automation involved.
  • Reads requests from in.txt and writes responses to out.txt.
  • Works with any MCP host that supports external servers.
  • Lightweight binary distributed via GitHub Releases.
  • Demonstrates the underlying MCP protocol in a transparent way.

Use cases of humanmcp

  • Learning the MCP protocol by manually handling requests and responses.
  • Testing MCP client behaviour with a predictable, human‑controlled server.
  • Fun / educational demonstration of how MCP servers communicate.
  • Exploring the structure of JSON-RPC messages in an MCP context.

FAQ from humanmcp

How does humanmcp differ from a normal MCP server?

A normal MCP server automates tool execution. humanmcp requires a human to read each request and manually write the response, making it the opposite of automation.

What dependencies or runtime are required?

Only the server binary (downloaded from Releases) and a text file editor/terminal. No language runtime, packages, or external services are needed.

Where are the request and response data stored?

All data is stored locally in two plain‑text files: in.txt (requests from the client) and out.txt (responses written by the human). The files are created by the MCP host.

Are there any known limits or disadvantages?

The server has no built‑in tools or logic – every response must be crafted manually. This makes it impractical for real‑world use and extremely slow compared to automated servers.

How does authentication and transport work?

The server uses standard MCP protocol over stdio (the binary reads from stdin and writes to stdout). humanmcp does not implement any authentication or custom transport; it simply relays the raw JSON-RPC messages via the text files.

コメント

「その他」の他のコンテンツ