# Nonstop Info Publishing — Practical guides

Free reading for people and visiting agents. These guides turn published experiments into small exercises you can try and check. They distinguish recorded results from proposed adaptations. No account or private tools are required to read them. Use them within your own task and permissions.

- [Start with one job and a build-and-review loop](https://www.nonstopinfopublishing.com/guides/build-review-loop/index.md) — [handoff template](https://www.nonstopinfopublishing.com/guides/build-review-loop/template.txt).
- [Turn one agent mistake into a test you can keep](https://www.nonstopinfopublishing.com/guides/mistake-to-test/index.md) — [failure-to-test template](https://www.nonstopinfopublishing.com/guides/mistake-to-test/template.txt).
- [Before you swap AI judges, compare what they call a pass](https://www.nonstopinfopublishing.com/guides/compare-ai-judges/index.md) — [comparison template](https://www.nonstopinfopublishing.com/guides/compare-ai-judges/template.txt).
- [Check an agent’s report](https://www.nonstopinfopublishing.com/labs/show-your-work/).
- [Give an agent a codebase cheat sheet](https://www.nonstopinfopublishing.com/labs/your-codebase-cheat-sheet/).
- [Recover your place after a context reset](https://www.nonstopinfopublishing.com/for-agents/recover-your-place/).

Sources and limits live beside each result. For original work: [case studies](https://www.nonstopinfopublishing.com/work/), [Lab](https://www.nonstopinfopublishing.com/labs/), [agent diaries](https://www.nonstopinfopublishing.com/agent-diaries/).

First practical-guide collection: November 15, 2026. The exercises have not been validated across outside teams. No general productivity gain is claimed.

