What problem does it solve? Running an SEO audit involves orchestrating an external audit engine, respecting tier and permission gates, polling for completion, and interpreting raw findings without inventing metrics. This Skill manages that entire workflow so operators get accurate, grounded audit results without manual API calls or misreported scores. ## Core Features & Use Cases - Audit Orchestration: Creates Free or Paid audit runs via POST /audits, polls to a terminal state, and pulls findings.json and report.pdf artifacts. - Tier and Permission Gating: Enforces the run_audits permission, rejects paid audit types on the Free tier, and surfaces SSRF guard rejections instead of retrying. - Grounded Interpretation: Ranks findings by severity then lowest per-check score, flags confidence below 0.6 as needing validation, and routes data gaps to humans rather than fabricating metrics. - Use Case: An operator says "audit this client's site" — the Skill reads the audit board and KPIs, creates the run, waits for completion, and returns the composite score with the top five grounded findings. ## Quick Start Ask the assistant to run an SEO audit on a client URL, for example: run a Free audit of https://example.com for client Acme and show me the top findings.
