meGPT - upload an author's content into an LLM

Name: meGPT - upload an author's content into an LLM
Author: adrianco

@adrianco

代码仓库

关于 meGPT - upload an author's content into an LLM

Code to process many kinds of content by an author into an MCP server

#megpt #content-processing #llm

基本信息

分类

AI 与智能体

许可证

Apache-2.0

运行时

python

传输方式

stdio

发布者

adrianco

配置

使用下面的配置,将此服务器添加到你的 MCP 客户端。

{
  "mcpServers": {
    "meGPT": {
      "command": "python",
      "args": [
        "mcp_server.py",
        "--author",
        "virtual_adrianco"
      ]
    }
  }
}

工具

未检测到工具

工具是从 README 中自动提取的。维护者可以在 ## Tools 标题下列出工具,即可填充这部分内容。

概览

What is meGPT?

meGPT is a repository and MCP (Model Context Protocol) server that organizes an author's public content—including books, blogs, podcasts, videos, and social media archives—into a searchable knowledge base for AI applications. It enables the creation of an LLM persona that can answer questions and generate summaries in the author's voice.

How to use meGPT?

Clone the repository, set up a Python virtual environment, and install dependencies. Run python build.py <author> to process content, then start the MCP server with python mcp_server.py --author <author> for STDIO mode or --transport streamable-http --port 8080 for HTTP.

Key features of meGPT

Supports books, blogs, Twitter archives, podcasts, and videos.
Includes an MCP server with semantic search and filtering.
Automatically detects and processes YouTube videos, playlists, and channels.
Handles Medium and Blogger archive extraction.
Generates MCP resources for any author directory.
Supports STDIO, HTTP, and SSE transports.

Use cases of meGPT

Building a chatbot that answers questions as if it were the author.
Summarizing an author’s opinions from decades of content.
Creating a searchable knowledge base for researchers or fans.
Training a RAG-based LLM persona from published work.

FAQ from meGPT

What content types does meGPT support?

It supports YouTube videos, podcasts, books (PDFs), stories, blog archives (Medium, Blogger), Twitter archives, presentation slides, and InfoQ videos.

How can I contribute my own content?

Clone the repo and add your own directory under authors/. If you want to contribute freely, submit a pull request.

What are the system dependencies?

The server requires Python 3 and packages listed in requirements.txt. Setup uses a virtual environment.

What license is the content under?

All content in this repo is Creative Commons Attribution-ShareAlike. Permission is explicitly granted for use as a training set for the meGPT concept.