Gsc Anomaly Detector

GSC Anomaly Detector

A [Claude Code](https://claude.ai/code) skill that turns your Google Search Console exports into a structured diagnostic brief — with ranked hypotheses, evidence confidence scores, and prioritised next actions.

Built for agency SEOs and in-house teams who need to diagnose GSC anomalies fast and communicate findings clearly to clients and stakeholders.

---

What It Does

Instead of describing what the data shows, this skill **diagnoses why it happened**.

It runs a structured 5-step framework on your GSC exports:

1. Scope — identifies the blast radius (pages, queries, device, country)
2. Timeline — maps the anomaly against algorithm updates and site changes
3. Hypothesis Stack — evaluates 12 ranked hypotheses with evidence signatures
4. Evidence Cross-Reference — for/against analysis with confidence scores
5. Client Brief — outputs a ranked hypothesis table with prioritised next actions

---

The 12 Hypotheses

#	Hypothesis	GSC Signal
H1	Algorithm update impact	Broad non-branded drop aligning to update dates
H2	SERP feature displacement (AI Overviews)	Position stable, CTR drops, impressions hold
H3	Canonicalisation issue	Page cluster impression drop, coverage excluded errors
H4	JS rendering / indexation failure	Dynamic content pages, gradual drop, coverage warnings
H5	Hreflang / locale targeting failure	Country-specific drop, locale traffic migration
H6	Redirect issue / broken URL	Page-level drop after deploy, 404s in coverage
H7	Internal link equity loss	Section-level drop after nav or structural change
H8	Content quality / E-E-A-T demotion	Evergreen content, gradual multi-week decline
H9	Mobile usability regression	Mobile-only drop in clicks and position
H10	Crawl budget / sitemap issue	Large site, broad impression drop, excluded spike
H11	GSC data anomaly / reporting lag	Sharp V-shape pattern that recovers