# CircuitProof

- **Event:** [Meta Global AI Developer Hackathon](https://cerebralvalley.ai/e/meta-ai-developer-hackathon)
- **When:** Oct 26, 12:00 PM – Nov 4, 9:00 PM (EST)
- **Where:** Online
- **Team:** [Chaya Chandana Doddaiggaluru Appajigowda](https://cerebralvalley.ai/u/Chaya)
- **GitHub:** https://github.com/chayachandana/circuitproof
- **Demo video:** https://youtu.be/AOHD4hhzzxc
- **Gallery:** https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery
- **Page:** https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/4419

CircuitProof uses Meta’s Muse Spark to investigate circuit faults, test repairs, and verify the results before applying a change.

Circuit debugging often involves trial and error: a suggested fix may sound plausible without actually restoring the intended behavior. I built CircuitProof to make each repair testable, with calculated evidence showing what changed and whether every target was met.

Users choose from five circuit examples or import a custom circuit with measurement targets. Muse Spark calls experiment tools to test component changes, inspect voltages and currents, and recommend a correction. The app accepts a repair only when it matches a passing experiment. After the user applies it, the browser independently recalculates the circuit and checks every target again.

The demo shows two simultaneous faults: a reversed LED and an undersized resistor. The agent tests a combined repair, and independent verification confirms that both LED currents return to the required range. Users can follow live investigation progress, inspect the experiment trace, compare before-and-after measurements, and export their notebook.

I built CircuitProof around the “Push the Frontier” theme. I wanted the AI to go beyond suggesting a fix to test it and show whether it works. Right now, the project verifies repairs in a simplified DC simulator. Testing those repairs on real hardware is the next step.

## More from Meta Global AI Developer Hackathon

- [Tellhouse](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/4416)
- [Newport1 team](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/4417)
- [Codeminister](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/4418)
- [styld.app](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/4420)
- [Acid](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/4421)
- [wb1936](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/4422)

---

Markdown version of https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/4419. Site index for agents: https://cerebralvalley.ai/llms.txt · full text: https://cerebralvalley.ai/llms-full.txt