Claude + Lean formally verifies Agent SDK: a few prompts yield 16 bug-fix PRs
shyamalanadkat · x · 2026-09-23
Meta engineer bcherny used Opus 5.5 with Lean to formally verify the Claude Agent SDK — a couple of short prompts produced 16 PRs fixing bugs and race conditions, with a demo video attached. He notes TLA+ also works well and sometimes combines the two to probe data flow, concurrency, and state management, adding that he doesn't know either language well himself but Claude excels at both.
The reblogger argues formal verification is the future of coding: the historical bottleneck — humans tediously specifying intent in languages like Lean — has been removed by frontier models.
More from coding & agent
- User automates entire video clipping workflow with Nous Research's Hermes agent — Teknium · 2026-11-12
- As models improve, writing good evals gets harder and more important — amankhan · 2026-11-12
- Prime agent ships with day-one multi-agent communication, runs in Minecraft unmodified — sarahcat21 · 2026-11-12
- Autonomous agent edits just 1,024 weights to lift Qwen3.5 0.8B from 7.4% to 37.4% on MATH-500 — VictorSanh · 2026-11-12
- Omarchy 4.6 lets you share local AI models across your tailnet, says DHH — alexcovo_eth · 2026-11-12
- Coding harnesses hand shell access to inference providers — 4-step gateway mitigation — shensi · 2026-11-12