Screen Recording to GitHub Issue

Say what you want on screen. Your coding agent turns it into one GitHub issue with numbered requirements, grounded in your code.

  1. 1Connect Claude Code (once per machine)
    claude mcp add --transport http --scope user vidmatic https://api.vidmatic.ai/mcp

    No key in it: run /mcp, choose Authenticate, and sign in from your browser.

  2. 2Record your screen, then hand your agent the recording
    /mcp__vidmatic__workstream_from_video <recording id>

    Use the id from your recording's link. Your agent reads it and files one GitHub issue.

Loading demo…
A change circled and talked through on screen, read by a coding agent with the code open, written up as an issue, built, and proven with before and after video.

Watch on its own page

How it works

1

Record the ask

Show the bug or the change and say what you want. Circle the parts that matter.

2

Your agent reads it with your code open

It reads the transcript and your marks, then finds the files they point at.

3

One grounded issue

It files exactly one GitHub issue: requirements R1 to Rn, each tied to what you said and to a file in your repo.

  • Requirements come from your words, not a summary of them
  • Every requirement points at real code
  • The recording stays linked to the issue

From screen recording to GitHub issue, step by step

Here is what happens between the moment you stop recording and the moment the issue opens in your repo.

What you record

Use the Vidmatic Chrome extension when position matters. Talk while you record, then pick up the pen when you reach the problem: circle the wrong button, cross out what should go, or write the right value by hand. Each mark is saved as an image at the moment you confirm it.

No extension? Press New Recording in the app and pick a screen or window. There is no pen in that mode, so say out loud which element you mean and pause on it for a second.

What your agent receives

When the recording finishes, choose Give to your agent and paste the prompt into Claude Code. Your agent calls get_recording_intel and gets the transcript in short timed segments, every mark with the moment you confirmed it and the sentence you were saying then, and a thumbnail. For each mark, get_annotation_image returns the screen with your mark on it. If the transcript is still processing, the agent waits the number of seconds the response gives and asks again.

What comes back

The agent searches your repository for where each ask lands and files one GitHub issue, however many asks the recording holds. Each ask gets a row with the time you said it, the mark you drew with it, your own words and one to three file and line anchors. When it cannot find where an ask lands, it writes "anchor: none found" instead of guessing. Then it starts the work, and you can follow each stage on Agent Workstreams.

When the issue already exists

If a teammate already filed an issue from a recording, run /mcp__vidmatic__issue_intel with the issue URL. Your agent reads the recording linked from the issue and posts the same table as one comment, so any ask the written ticket dropped comes back.

Questions

How do I turn a screen recording into a GitHub issue?
Connect Claude Code to Vidmatic, record your screen, then run /mcp__vidmatic__workstream_from_video with the recording id. Your agent reads the recording and files one issue in your repo.
Who writes the issue, Vidmatic or my agent?
Your agent does, in your repo, with your code open. Vidmatic gives it the recording's transcript, marks and frames.
What if someone else recorded it?
A recording sent to you carries a watch token. Your agent passes it on every call and files the issue itself.
Which tracker does the issue go to?
GitHub only. This loop files GitHub issues, in your repo, through your own agent.
  • Screen Recording for AI Coding Agents

    Screen recording for AI coding agents: record your screen, circle it and say it. Claude Code or Cursor reads the transcript, marks and frames over MCP.

  • AI Agent Proof-of-Work Demos

    Proof of work for AI coding agents: when your agent says done, Claude Code records your running app and narrates the change, so you can see it before you merge.

  • Create a Ticket From a Video

    Create a ticket from a video. Record your screen, add it to Agent Workstreams, and your dispatcher files a GitHub issue in the right repo.

Turn Your Next Recording Into an Issue

Connect Claude Code once, then hand it a recording.

Explore more Vidmatic use cases

Docs: Run workstream_from_video