Post

LLMs write code fast, but how do you trust it? Unit tests only cover paths you think to test. Bounded model checking exhaustively explores all paths and gives mathematical proofs when it finds bugs. 🧵

I built a Claude Code plugin for ESBMC (esbmc.github.io) — bringing formal verification into your agentic coding workflow. Ask Claude to check your code, ESBMC runs under the hood, and you get concrete counterexample traces when it finds bugs.

0 reposts2 likes