# yt-mem-ai

A local memory for YouTube: transcripts, search across them, summaries and timestamped highlights. Everything runs on your machine.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed, Gemini CLI · ★ 7
- Author: dasein108
- License: MIT · policy=link-only · description only; the file stays with the author
- Source: https://github.com/dasein108/yt-mem-ai
- Updated: 2026-08-23
- HTML: https://affpapa.org/en/nevibecode/mcp/yt-mem-ai

## What it does

The server lets your assistant work with videos you care about. It pulls YouTube captions when they exist and falls back to offline Whisper when they don't. Every transcript is kept on disk and indexed, so the library stays searchable. Ask what a video said about a topic and you get an answer pointing at the moment.

Writing is left to the model you already have connected: summaries, highlights, Q&A, slide decks, in the language of the video itself. Subscriptions are handled too — new uploads land in a single daily digest. Mark videos you liked or didn't, and recommendations come from your own library.

Two ways in. An interactive wizard asks what you want and which apps to wire up. Or add one entry to your MCP config: `uvx` fetches the package on the host's first launch and caches it after that. That first start takes longer than the host's startup check allows, so the tools may not appear. Run `uvx --from 'yt-mem-ai[mcp]' yt-ai-mcp --help` once and reopen the app. GUI apps often miss `~/.local/bin` on `PATH`; absolute paths fix that.

## Requirements
- Needs: Python and uv (commands run via uvx)

## Install

### Claude Code (native platform)

**Platform CLI**
```bash
claude mcp add yt-mem-ai -- uvx yt-mem-ai
```
uvx runs the command named after the package; if the README names another one, use uvx --from yt-mem-ai <command>

### Codex

**Platform CLI**
```bash
codex mcp add yt-mem-ai -- uvx yt-mem-ai
```
uvx runs the command named after the package; if the README names another one, use uvx --from yt-mem-ai <command>

### Cursor

**Config file**
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "yt-mem-ai": {
            "command": "uvx",
            "args": [
                "yt-mem-ai"
            ]
        }
    }
}
```
uvx runs the command named after the package; if the README names another one, use uvx --from yt-mem-ai <command>

### GitHub Copilot

**Config file**
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "yt-mem-ai": {
            "type": "stdio",
            "command": "uvx",
            "args": [
                "yt-mem-ai"
            ]
        }
    }
}
```
uvx runs the command named after the package; if the README names another one, use uvx --from yt-mem-ai <command>

### Gemini CLI

**Platform CLI**
```bash
gemini mcp add yt-mem-ai uvx yt-mem-ai
```
uvx runs the command named after the package; if the README names another one, use uvx --from yt-mem-ai <command>

## When to use it
- transcribe a video that has no captions
- jump to the moment a topic came up in a talk
- turn a day of new uploads into one digest
- build slides or a PDF from a video's content
- search your own watch history without a cloud service

## Example prompts
- Summarize this video: `https://youtu.be/…`
- Find where my library covers transcript indexing
- Build a digest from my subscriptions' new uploads

## Similar
- [REA](https://affpapa.org/en/nevibecode/mcp/rea.md) — MCP server · ★ 386989
- [GitHub](https://affpapa.org/en/nevibecode/mcp/github.md) — MCP server · ★ 33918
- [Netdata](https://affpapa.org/en/nevibecode/mcp/netdata.md) — MCP server · ★ 81115
- [browser-use](https://affpapa.org/en/nevibecode/mcp/browser-use.md) — MCP server · ★ 119160
- [Codebase Memory](https://affpapa.org/en/nevibecode/mcp/codebase-memory.md) — MCP server · ★ 47392
- [mcp-server](https://affpapa.org/en/nevibecode/mcp/mcp-server.md) — MCP server · ★ 43872

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md