gstack /review catches 22 of 84 blocking bugs vs 3 for plain Opus, evals show
garrytan · x · 2026-11-05
Garry Tan endorses gstack's /review skill as genuinely useful for preventing bugs. Per the attached eval data, on Opus 5.5 gstack's /review caught 22 of 84 known blocking bugs versus 3 for plain Opus — a +22.6 point gain (95% CI [+11.0, +33.7]). The numbers come from gstack-evals, and the author's team reproduced every figure from scratch on a clean machine without keys, calling it the strongest pro-gstack evidence available.
More from coding & agent
- GitHub Next's PDD Kit uses formal methods to prove AI-generated web app code is correct — 0xkarasy · 2026-11-05
- Meta's JiTTesting deploys LLM-generated tests to catch bugs pre-merge, cutting human review 70% — AxSaucedo · 2026-11-05
- Built an app in a weekend with AI, now spending 3 weeks figuring out what it does — kylegawley · 2026-11-05
- Developer seeks open-source guardrails against malicious file uploads to LLM chatbots — SnooCupcakes982 · 2026-11-05
- Dev predicts 180Hz screen-frame models will make vibe coders obsolete within a year — ryunuck · 2026-11-05
- Agents ace isolated tasks but drop to ~38% at maintaining codebases over time — FullStackFrenzy · 2026-11-05