I built an open-source tool that records screen + voice bug reports producing AI-ready markdown

I kept running into the same problem in AI-assisted dev workflows: bug reporting was still manual and slow.

You find an issue, then switch to writing mode — describe context, take screenshots, organize everything, and rewrite it for your coding agent.

So I built markupR.

It takes screen + voice and turns it into structured markdown with inline screenshots that AI tools can use directly.

Workflow

1. Record your screen while narrating what’s wrong
2. Transcription + timestamp analysis
3. Frame extraction at relevant moments
4. Output: AI-ready markdown report

Why I’m building it this way

• Open source (MIT)
• Free forever core workflow
• BYOK if you want cloud models
• Local options if you don’t

I’m not trying to build lock-in here — I want a tool people actually use and improve over time.

If you try it, I’d love feedback on:

• where your current bug-report handoff breaks
• whether this output format is useful in real projects
• what would make this production-worthy for teams

Repo + demo: Github
(If useful, I can also share architecture details and implementation notes.)