seo-prune-auditlisted

# SEO Prune Audit Skill Goal: tie a set of pages to three data-backed decisions — **(1) which should be closed/merged, (2) which should be kept, (3) which deserve authority investment (backlinks/content)** — and deliver it as a management-ready Excel workbook plus a summary. ## Prerequisites - Access to a GSC BigQuery bulk export (see `config.json` → `bq.table`). - Ahrefs MCP connected (for the authority module). Check the unit quota with `subscription-info-limits-and-usage` (a free call) before running. - Python 3 + openpyxl (for local computation and the report). - Queries run through `scripts/bq_run.ps1` (REST API). If the `bq` CLI works in your environment you can adapt the SQL to it instead. ## Data sufficiency — read first Closure/merge decisions are only trustworthy with enough history. This skill is built for a **GSC BigQuery bulk export covering ideally ~12 months** (at least `thresholds.min_history_days`, default 180 days). If the user only provides a **short window** — a 28-day or 3-month export, or a copy/paste from the Search Console UI — **do not recommend closing, merging, or noindexing anything.** Such pages are automatically held as `REVIEW` (not put in any rollout batch), and step 04 prints a warning. In that case, tell the user the data is too thin for removal decisions and ask for a longer export before acting; you may still describe traffic patterns and authority opportunities, which do not carry the same risk. A page with few impressions over 2