SpudStack uses Grok to review agent plans with PASS/FAIL verdicts — agents can't sign off their own work
alexcovo_eth · x · 2026-10-20
The author built SpudStack, a Grok-powered bot that reviews AI agents' plans using the pstack method: it demands re-runnable proof, never lets an agent approve its own work, and issues only PASS, FAIL or PASS-WITH-NOTES verdicts with specific fixes. pstack iterates fast (0.15.26 → 0.15.28 overnight, auto-updated locally), adds a maintain-formal-verification-skill that keeps formal proofs in sync after code changes (TLA+ by default with fallback), offers formal verification setup in setup-pstack, and lifts sharpen's ban on rewriting prompts.
Related event: SpudStack Launches on Grok Bot Marketplace as Strict Agent Plan Reviewer(2 posts)→
More from coding & agent
- CDSE Paper Splits Agent Skills Into Four Components, Gains 7.6 Points — ruslansv · 2026-10-20
- Microsoft open-sources Scope to evaluate how AI coding agents use your SDKs — DanWahlin · 2026-10-20
- Descript adds one-click import for Claude Motion animations, turning them into narrated videos — descript · 2026-10-20
- Dev builds a game demo in 2 days with Opus 5.5, Blender and Tripo assets — techartist_ · 2026-10-20
- Even Opus 5.5 hasn't internalized the bitter lesson when writing LLM systems — brandon_galang · 2026-10-20
- After 400+ agents, an engineer maps 15 AI agent memory patterns from types to control — MaryamMiradi · 2026-10-20