# talkthru

Feeds narrated screen recordings to coding agents by pinning speech to frames. Runs locally with whisper.cpp and requires zero API keys.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed, Gemini CLI · ★ 5
- Author: EdonZo
- License: MIT · policy=link-only · description only; the file stays with the author
- Source: https://github.com/EdonZo/talkthru
- Updated: 2026-08-25
- HTML: https://affpapa.org/en/nevibecode/mcp/talkthru

## What it does

talkthru converts voiceover screen recordings into actionable agent context. You tap through your mobile or desktop UI while explaining bugs out loud. The tool segments your video into frames and attaches transcribed remarks to the exact moment you spoke them.

Processing stays entirely on your computer. Running `npx talkthru doctor --fix` fetches `ffmpeg` alongside `whisper.cpp`. The watcher `npx talkthru watch` monitors your Downloads folder and reacts only to matching video names.

Make sure to mute audio coming from the application itself. The speech model cannot distinguish system sounds from your voice, which corrupts the transcript. Forgetting to enable the mic will result in an empty review session.

## Requirements
- Needs: Node.js (commands run via npx)

## Install

### Claude Code (native platform)

**Platform CLI**
```bash
claude mcp add talkthru -- npx -y talkthru-mcp
```

### Codex

**Platform CLI**
```bash
codex mcp add talkthru -- npx -y talkthru-mcp
```

### Cursor

**Config file**
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "talkthru": {
            "command": "npx",
            "args": [
                "-y",
                "talkthru-mcp"
            ]
        }
    }
}
```

### GitHub Copilot

**Config file**
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "talkthru": {
            "type": "stdio",
            "command": "npx",
            "args": [
                "-y",
                "talkthru-mcp"
            ]
        }
    }
}
```

### Gemini CLI

**Platform CLI**
```bash
gemini mcp add talkthru npx -y talkthru-mcp
```

## When to use it
- feed spoken bug walkthroughs directly to coding agents
- pin user feedback to exact UI video frames
- extract UI component selectors via an optional hierarchy JSON file
- diagnose interface issues without manual screenshot cropping

## Example prompts
- Inspect the feedback from my latest talkthru recording and patch the bugs
- Check the last talkthru session and resize the checkout button
- Review the recorded screens and speech to fix the missing error message

## Similar
- [REA](https://affpapa.org/en/nevibecode/mcp/rea.md) — MCP server · ★ 392262
- [GitHub](https://affpapa.org/en/nevibecode/mcp/github.md) — MCP server · ★ 33989
- [Netdata](https://affpapa.org/en/nevibecode/mcp/netdata.md) — MCP server · ★ 81163
- [browser-use](https://affpapa.org/en/nevibecode/mcp/browser-use.md) — MCP server · ★ 119231
- [Codebase Memory](https://affpapa.org/en/nevibecode/mcp/codebase-memory.md) — MCP server · ★ 47577
- [mcp-server](https://affpapa.org/en/nevibecode/mcp/mcp-server.md) — MCP server · ★ 43902

---
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