---
title: "Improve"
description: "Start with the question you need to answer. Use the same cases when comparing a change, and keep failures as regression cases."
canonical_url: "https://docs.oz0.ai/improve"
---
# Improve

> Start with the question you need to answer. Use the same cases when comparing a change, and keep failures as regression cases.

For: People testing and improving agents.

Start with the question you need to answer. Use the same cases when comparing a change, and keep failures as regression cases.

- [Why did this job fail?](https://docs.oz0.ai/improve/errors-and-insights): Find recurring failures and open the affected jobs.
- [What changed?](https://docs.oz0.ai/improve/search-and-compare): Compare jobs and agent versions.
- [Where should I fix it?](https://docs.oz0.ai/improve/where-a-fix-belongs): Choose the part that can correct the failure at the lowest recurring cost.
- [Does the change pass?](https://docs.oz0.ai/improve/evals): Write eval cases and gate plugin updates.
- [Is the answer good enough?](https://docs.oz0.ai/improve/outcome-checks): Configure judges, reviewers and candidate selection.
- [Is the new definition better?](https://docs.oz0.ai/improve/comparisons-of-a-changed-agent): Compare quality on paired cases.
- [Does it cost less?](https://docs.oz0.ai/improve/measured-savings): Measure billed cost per successful run.
- [How do the starters perform?](https://docs.oz0.ai/improve/measured-results): Run and inspect the starter agents' evals.
- [Can an agent inspect the evidence?](https://docs.oz0.ai/improve/evidence-for-agents): Grant bounded read access to results.


## Sitemap

See the full [sitemap](https://docs.oz0.ai/sitemap.md) for all pages.