seo-google

Trigger phrases

Phrases that activate this skill when typed to Claude Code:

• search console
• GSC
• PageSpeed
• CrUX
• impressions

What it does

seo-google is a Claude Code skill from AgriciDaniel’s claude-seo repo. It bridges the gap between static analysis and Google’s real-time field data: actual Chrome user Core Web Vitals, real indexation status per URL, search performance (clicks, impressions, CTR, position), and organic traffic trends. All APIs are free — they require a Google Cloud project with an API key and/or service account, but cost nothing per call.

The skill is organized into four credential tiers: Tier 0 (API key only) unlocks PageSpeed Insights, CrUX, CrUX 25-week history, YouTube search, and NLP entity analysis; Tier 1 (service account) adds Search Console, URL Inspection, sitemap status, and Indexing API; Tier 2 adds GA4 organic traffic; Tier 3 (Google Ads account) adds Keyword Planner volume. The skill detects your credential tier at startup and communicates it before running any command. It also offers PDF/HTML report generation via /seo google report after any analysis.

When to use it

Reach for it when:

• You need real CrUX field data rather than Lighthouse lab scores — field data reflects actual users’ experience
• You want to verify whether a URL is actually indexed by Google (URL Inspection returns the real verdict, not an inferred one)
• You need Search Console click/impression data to find quick-win keywords at positions 4–10

When not to reach for it:

• Google API credentials are not configured — run /seo google setup first; the skill will walk you through the setup step by step
• You need CrUX data for a low-traffic site; CrUX requires sufficient Chrome user volume and returns a 404-equivalent when data is insufficient (not an auth error — the skill distinguishes these)

Install

Copy the seo-google SKILL.md into .claude/skills/seo-google/ along with the scripts/ directory from the repo root.

Trigger phrases: “search console”, “GSC”, “PageSpeed”, “CrUX”, “field data”, “indexing API”, “GA4 organic”, “URL inspection”, “real CWV data”, “impressions”, “clicks”, “CTR”, “LCP”, “INP”, “CLS”.

Run /seo google setup to configure credentials. Commands are tier-gated: the skill tells you which tier is active before every command.

What a session looks like

A typical session has three phases:

1. Credential detection. The skill runs python scripts/google_auth.py --check --json and reports which tier is active, which commands are available, and what to configure to unlock the next tier. This takes a few seconds and runs before any other command.
2. Data retrieval. Commands follow the pattern of fetching one data type at a time: pagespeed <url> for lab + field CWV, gsc <property> for search performance, inspect <url> for indexation status. Each uses a dedicated script with JSON output.
3. Report generation. After collecting data, /seo google report <type> generates a professional PDF with charts: CWV gauges and 25-week trend lines for cwv-audit, query tables with quick-win highlights for gsc-performance, and coverage donut charts for indexation reports.

Receipts

Works well: The crux-history <url> command showing 25 weeks of Core Web Vitals field data is genuinely difficult to replicate without this setup — it shows whether your CWV improvements are actually landing for real users or just in lab conditions. The quick-win detection in GSC (queries at positions 4–10 with high impressions) is a reliable first pass for finding content to optimize.

Backfires: The Indexing API is officially supported only for JobPosting and BroadcastEvent/VideoObject pages. The skill documents this restriction clearly and informs the user — but some people expect it to force-index any URL, which it does not reliably do for general content.

Pattern that works: Use inspect-batch after any migration that touches URL structure or canonicals. Batch inspection of your top 50 pages takes a few minutes and tells you the real indexation verdict for each — it catches canonicalization mistakes that won’t show up in GA4 for two to three weeks.

Source and attribution

Originally written by AgriciDaniel. The canonical SKILL.md and supporting scripts live in the seo-google folder of the claude-seo repository.

License: MIT. Install, adapt, and redistribute with attribution preserved.

This page documents the skill from a practitioner’s perspective. For the formal spec and updates, defer to the source repo.