policy-monitor

Privacy policy drift occurs when real-world data practices diverge from configured policy commitments. You detect it by comparing saved privacy artifacts and proposed practices against authoritative policy text to identify covered, missing, or conflicting gaps.

To check if a proposed data practice is covered, run a direct query diff against your current policy commitments. This classifies the practice as covered, missing, or conflicting, and drafts suggested policy language with update timing guidance.

Conduct a compliance review by running a sweep-based triage of saved PIAs, DPA reviews, and triage outcomes. The sweep extracts approved practices and diffs them against your current policy to classify gaps as REQUIRED versus ADVISABLE.

Yes, multi-surface consistency checks review multiple policy commitment surfaces beyond the web policy. This includes scanning cookie/CMP promises, app store labels, in-product consent flows, and sector notices to identify conflicting gaps.

You need a configured CLAUDE.md state file, a located policy document, and an outputs folder. The drift detection process reads this state to locate your policy text and saved privacy artifacts before producing gap classifications.

REQUIRED and ADVISABLE are gap classifications used when policy drift is detected. REQUIRED gaps indicate mandatory policy updates needed for compliance, while ADVISABLE gaps suggest recommended language updates with specific timing guidance.