What problem does it solve? Manually reviewing a backlink profile for spammy or toxic links is slow and error-prone, and flagging links for disavow without a grounded, threshold-based process risks either missing real toxicity or over-flagging healthy links. ## Core Features & Use Cases - Grounded Profile Review: Reads real backlink rows (authority, spam, anchor, referring domain) and KPIs from the AIOS API without inventing DA/DR or spam numbers. - Anchor Toxicity Assessment: Analyzes the anchor distribution (branded vs exact-match vs naked/generic) against the Team C rubric, surfacing over-optimization and PBN/link-farm footprints. - Guarded Toxic Flagging: Proposes a spam threshold, stops for explicit LEAD confirmation, then flags at/above-threshold links as toxic to queue a disavow review (LEAD-only write). - Use Case: An operator asks for a "backlink audit" on a client; the skill reads the profile, recommends a spam threshold of 60, shows which 12 referring domains would be flagged, and after the LEAD confirms, queues them for disavow review. ## Quick Start Ask the assistant to run a backlink audit for a client and propose a spam threshold for flagging toxic links.
