consolidate-sensors

Consolidating runtime verdicts into per-sensor append-only notes requires distilling evidence from cycle verdicts and progress data, then appending bullets with citations to individual sensor files. This process reads runtime files to compute diffs and trigger frontmatter recalibration.

Sensor calibration maintenance based on runtime evidence is the process of distilling active task progress and verdicts into auditable, append-only per-sensor records. It tracks known issues and frequent errors to improve sensor reliability across cycles via frontmatter recalibration.

Automating per-sensor distillation involves reading runtime verdicts and progress data to generate cited bullets for each sensor. The automation computes diffs for known issues, frequent errors, and calibration sections, appending history to sensor bodies without overwriting previous records.

Triggering sensor frontmatter recalibration requires access to `.yoke/runtime/.judge-verdicts/`, `.yoke/runtime/progress.md`, and `.yoke/config.yaml`. These files provide the cycle evidence and configuration necessary to compute diffs and detect calibration drift.

Maintaining an append-only audit trail for sensor calibration is best achieved by distilling cycle verdicts into per-sensor files with mandatory citations. This approach preserves historical evidence by appending bullets and recalibrating frontmatter costs only when drift is detected.

Using append-only records for runtime sensor evidence preserves an auditable history of known issues and frequent errors across cycles. This governance model prevents data loss during distillation, enabling accurate diffs and reliable frontmatter recalibration for sensor maintenance.