Researcher Uses AI and Formal Verification to Unearth Three NetHack Bugs
davidbau · x · 2026-09-27
David Bau shares his hands-on workflow learning formal methods coding with AI assistance: he found three reproducible bugs in NetHack's polymorph code (src/polyself.c) — a deleted light source that was never created, cleanup executed twice per polymorph causing double damage rolls, and decisions still made by the superseded form's rules (fatally stripping water-walking boots while standing in lava).
Formal verification then revealed a hole in his first fix: a second form change triggered by polymorph damage is mishandled. He filed issue #1682 with a companion PR and asks how to avoid missing issues when checking assumptions.
Related event: MIT Researcher Uses AI and Formal Verification to Uncover NetHack Bugs(3 posts)→
More from coding & agent
- Baseten's Guide: Evaluate Open Models for Coding Agents on Your Own PRs — baseten · 2026-11-16
- Google Research open-sources PaperOrchestra, a multi-agent framework for writing AI papers — google-research · 2026-11-16
- ffmpeg-skill gives coding agents local video editing via 42 FFmpeg tools and MCP — kajisho5 · 2026-11-16
- Agents Now Reply to Slack Messages on Their Own, Resolving a Prod Incident Ghost-to-Ghost — iamrobotbear · 2026-11-16
- Engineer's Interview Prep Method: Use Claude to Find Which Answers Undersell Your AI Work — brandon_galang · 2026-11-16
- DeepSeek open-sources its app frontend "chat-demo", already at 3.9k stars — NielsRogge · 2026-11-16