# Factory assessment

Coordinators post this on their PR with `add_issue_comment` right after opening it, so a reviewer can pick a review depth before reading the diff. It complements the PR body, which describes the change but never grades it. Keep each line to one sentence, with nothing outside the template. Post it once per PR. When a later change makes a line wrong, edit that comment with `update_issue_comment`; never post a second assessment.

```markdown
## Factory assessment

**Fit:** <implements the issue fully | partially, and what is left>
**Evidence:** <the reproducer's or analyst's failing test and acceptance test before the change and its rerun of both after, or why they could not run, and the commands that passed after the change>
**Risk:** side effects <none/low/medium/high>, performance <…>, backwards compatibility <…>, from the area reviewers; security <…>, from the security reviewer (finish lane only)
**Review depth:** <glance | focused | deep>, because <reason>
```

Review depth:

- **glance**: only docs, tests or comments change.
- **deep**: the change touches a public API, a GraphQL or OpenAPI schema, a migration, auth or permissions, or any risk is medium or high.
- **focused**: everything else.