last-call

An end-of-session review identifies your session scope by finding commits and uncommitted changes, reads every changed file end-to-end, and generates a structured report rating findings by severity to surface bugs and architecture gaps.

Reviewing code changes for documentation debt requires reading every changed file end-to-end with production-grade rigor, evaluating correctness and architecture, and producing a structured report with recommended actions and severity ratings.

Mechanical fixes for artifacts like typos and debug prints are applied automatically during the review process, but structural code modifications are strictly isolated and require explicit user confirmation before execution.

Yes, you can generate a structured quality report by identifying the session scope which includes session commits, uncommitted changes, and stashed work, then reading those changed files end-to-end to evaluate correctness.

Preparing for a project handoff involves performing a deep review of all changed files to produce a deterministic, action-oriented report that rates findings as critical, warning, or nit, ensuring thorough close-out and milestone verification.

You should avoid using an automated session-end review if you expect the tool to modify code logic autonomously, because it isolates fixes to mechanical artifacts and strictly requires user confirmation before applying any non-mechanical code changes.