# FirmReverse

- **Event:** [Gemini 4 Argon SF Hackathon](https://cerebralvalley.ai/e/gemini-4-argon-sf)
- **When:** Sun, Oct 25 at 9:00 AM – 10:00 PM (PDT)
- **Where:** San Francisco, CA
- **Team:** [Sissi Wang](https://cerebralvalley.ai/u/sissississi_013), [Iana Lin](https://cerebralvalley.ai/u/ianalin)
- **GitHub:** https://github.com/ianalin123/firm-reverse
- **Demo video:** https://www.loom.com/share/90a3efd9a8eb406b99c934dc46ad343c
- **Gallery:** https://cerebralvalley.ai/e/gemini-4-argon-sf/hackathon/gallery
- **Page:** https://cerebralvalley.ai/e/gemini-4-argon-sf/hackathon/gallery/69

Reverse engineering hardware usually takes about $10k of lab gear. We think a phone, a logic analyzer and a model that checks its own work get you most of the way. Firm 

Reverse runs in three stages on Gemini 4 Argon. Each stage is hypothesize, verify, revise. 

First, a camera guide. Argon makes claims about the device and cites the shot numbers behind each one. When a new shot contradicts a claim, it revises it and records why. Then it asks for the shot that settles the weakest claim. A datasheet agent cross checks every chip reading. Second, a required probe step. The guide tells you which Saleae channel goes on which pin. You upload the capture and we decode it. That becomes the strongest evidence. Third, a verified 3D reconstruction. Argon builds the device in Three.js, renders it, compares it to the photos and patches it until it matches. It ends with a clean room functional spec. It never copies firmware.

## More from Gemini 4 Argon SF Hackathon

- [My Little Brain](https://cerebralvalley.ai/e/gemini-4-argon-sf/hackathon/gallery/66)
- [polar bears](https://cerebralvalley.ai/e/gemini-4-argon-sf/hackathon/gallery/67)
- [Footage-to-World](https://cerebralvalley.ai/e/gemini-4-argon-sf/hackathon/gallery/68)
- [SatMission](https://cerebralvalley.ai/e/gemini-4-argon-sf/hackathon/gallery/70)
- [Mise](https://cerebralvalley.ai/e/gemini-4-argon-sf/hackathon/gallery/71)
- [Scailer](https://cerebralvalley.ai/e/gemini-4-argon-sf/hackathon/gallery/72)

---

Markdown version of https://cerebralvalley.ai/e/gemini-4-argon-sf/hackathon/gallery/69. Site index for agents: https://cerebralvalley.ai/llms.txt · full text: https://cerebralvalley.ai/llms-full.txt