Happy slice of mcp pizzaMCP.pizza

talkthrough-mcp

MCP.Pizza Chef: korovin-aa97

Talkthrough-mcp lets your AI assistant transform your screen recordings into detailed transcripts, key video frames, text extracted from images, and precise timestamps. You can use it to create clear, evidence-backed bug reports or review meetings and demos with searchable video content. It works locally on your computer with apps like Claude Desktop, Cursor, and VS Code, requiring no internet connection or API keys. Setup involves just adding a small configuration snippet to your existing app.

Coding
Other

Use This MCP server To

Turn a screen recording into a searchable transcript Extract text from video frames using OCR Create a timeline of key moments in a video Generate a draft bug report from a narrated recording Review meeting recordings with searchable transcripts Analyze product demos with time-stamped evidence Investigate production incidents from video logs

README

talkthrough-mcp

ci license: MIT python PyPI

Quickstart · Tools · Benchmarks · FAQ · Troubleshooting · Changelog · Contributing

Don't write a bug report. Record it.

Give Claude Code or Codex a narrated .mov/.mp4 — talkthrough turns it into searchable transcript, exact frames, OCR and wall-clock timestamps, locally — so your agent writes an evidence-backed issue draft or investigates the fix.

Also works for meetings, workshops, product demos, and production incidents.

Illustration of a /talkthrough:bug run: a recorded checkout bug is indexed locally, the evidence is found, and a ready-to-file issue draft is assembled

Illustration — an animation of the /talkthrough:bug storyline, not a screen capture: the recording is indexed locally (transcript · keyframes · OCR · wall-clock), the evidence checkpoint is assembled, and the draft is ready to file with your own tracker tooling (gh, Jira, a GitHub MCP server — talkthrough itself never leaves your machine and never posts anything).

Real runs, no animation:

  • ▶ Watch the demo with sound (1:18) — an unedited session: a narrated recording goes in, a ready-to-file bug-report.md comes out.
  • Silent recording → issue draft — the whole thing as files you can re-run: a Playwright-recorded, audio-free .mp4 and the unedited agent output. Every number is reproducible — processing that file gives job_id 8703a66bbe77a7d0 (the job id is the sha256 prefix, so shasum -a 256 predicts it), 17 keyframes / 4 unique, and 0 transcript segments because there is no audio track.

Quickstart

One command, no system dependencies: ffmpeg falls back to a bundled build, OCR is pip-only, and whisper models download themselves on first use. The only prerequisite is uv (brew install uv or curl -LsSf https://astral.sh/uv/install.sh | sh).

talkthrough-mcp FAQ

Can I use talkthrough-mcp to create a searchable transcript from a screen recording?
Yes — it converts your video into a transcript, key frames, and timestamps locally for easy searching.
Can I use this to generate a bug report draft from a narrated video?
Yes — it processes your recording and helps your AI draft an evidence-backed issue report.
Which apps support talkthrough-mcp?
It works with Claude Desktop, Cursor, VS Code, Cline, Goose, Windsurf, Roo Code, and more.
Do I need an API key or account to use talkthrough-mcp?
No — it runs locally and does not require any API keys or accounts.
How hard is it to set up talkthrough-mcp?
Setup is copy-paste-config — you add a small config snippet to your existing app.
Does talkthrough-mcp send my recordings to the cloud?
No — all processing happens locally on your machine, keeping your data private.