introspection-review-tracker

You persist code review findings by saving them as self_assessment memories. This tracks review outcomes automatically after adversarial code reviews to enable continuous learning and auditability.

The best way to track adversarial review outcomes is using an automated hook that maps severity tiers to error categories. This enforces consistent tracking and persists findings directly to your memory store.

Severity-to-category mapping translates review severity tiers into standardized error categories. This enforces a consistent content format for self_assessment memories stored during introspection workflows.

Yes, you need the llmem track-review CLI to persist findings. You also need a valid review findings JSON and access to the memory store to complete the self_assessment memory creation.

You need to persist review findings as memories after every adversarial code review. This automated post-review verification step ensures findings are tracked for future introspection and auditability.

The review findings JSON must conform to the severity-to-category mapping content format. This structured format ensures self_assessment memories are created correctly in the memory store by the llmem CLI.