poteto ships two skills that teach coding agents to verify their own work
kieranklaassen · x · 2026-10-12
poteto (author of pstack) argues the most basic agentic loop is verification: letting agents run, control, and debug the app like a real user so they can close the loop themselves.
He added two skills to pstack that anyone should use or copy:
- /create-verification-skill: generates a skill teaching your agent how to run, control, and debug your app, plus a feature map — a user-POV map of every feature and how to reach it, letting agents navigate the app like a real user and greatly improving self-verification.
- /maintain-verification-skill: run as a daily automation to keep the feature map fresh, since it goes stale quickly.
The goal: agents produce an artifact proving the work is correct, matches intent, and is easier to review than reading the whole PR. For state-machine bugs, agents can even model the problem in TLA+.
More from coding & agent
- TestSuiteBench: coding agents kill only ~55% of bugs with generated test suites — lbeurerkellner · 2026-10-13
- Running coding agents in YOLO mode: a practical 5-layer guardrail setup — Arindam_1729 · 2026-10-13
- Cloudflare Agents SDK adds web_search (Exa/Linkup/ceramic) and Markdown web_fetch tools — irvinebroque · 2026-10-13
- 45 minutes with AI: failure cases, BQ datasets, and a generative UI — rseroter · 2026-10-13
- DeepSeek desktop agent app reviewed: model promising, harness trails Cursor in 5 ways — casper_hansen_ · 2026-10-13
- Hone, founded by ex-Cognition and Mercor staff, builds business-running agents — nmasc_ · 2026-10-13