Google SEO Data Assistant
Pull first-party Google data for search performance, indexing, Core Web Vitals, and organic traffic analysis.
The skill documents credential tiers, OAuth scopes, service-account permissions, quotas, and the official restriction of the Indexing API, while describing local credential storage and security controls. It does not clearly require per-operation confirmation for write actions, batch indexing can create external side effects, and data-flow, rollback, and report-cleanup guidance is incomplete, so points are deducted.
Commands, scripts, failure scenarios, and retry guidance are fairly explicit, and repository context includes CI and test evidence. However, this review is static and did not reproduce seo-google key paths; runtime availability, dependency installation, abnormal-output behavior, and edge cases remain unverified. Potential documentation inconsistencies such as batch limits also reduce the score.
Trigger phrases, user-invocable status, argument hints, credential tiers, and primary scenarios are clear across GSC, CWV, GA4, indexing, and keyword analysis. Boundaries, non-fit cases, Chinese-language support, and mainland-China network reachability are insufficiently specified, while core functions depend on Google services that may be inaccessible, so points are deducted.
The selected document provides progressive sections, command tables, prerequisites, references, output conventions, and error handling. Repository context supplies MIT licensing, versioning, authorship, CI, security policy, and maintenance signals, but the skill itself lacks a clear changelog, maintenance owner, and update path; some dependencies and output contracts require consulting other files, so points are deducted.
The skill targets practical Google SEO data collection and defines usable metric, JSON, and report outputs, offering convenience over manual querying. Static evidence does not verify representative outputs or end-to-end completion, and several features require multi-tier credentials, quotas, and complex report dependencies, so the score is conservatively capped and reduced.
API references, quota tables, authentication procedures, error handling, and repository CI/test material provide some auditability. There are no committed, key-path execution results or independent corroboration specific to this skill, and no scripts were executed in this review, so only limited credit is warranted.
- Add explicit per-run confirmation, URL scoping, and reversibility guidance before executing /seo google index or index-batch; misuse can create unnecessary external notifications.
- Google API, GA4, GSC, and Google Ads features depend on external accounts, permissions, quotas, and network reachability; mainland-China users may not be able to use core functions reliably.
- Reconcile batch-indexing limits, freshness statements, and quota values across the implementation and references, and add reproducible tests for the key commands.
- Service-account JSON, API keys, and GA4/GSC data are sensitive; verify local file permissions, log redaction, and report-output storage boundaries before use.
What does this skill do, and when should you use it?
This is one focused skill inside the claude-seo repository, not the repository’s full SEO audit system. It connects to Search Console, PageSpeed Insights, CrUX, the Indexing API, GA4, and selected YouTube, Natural Language, Knowledge Graph, and Web Risk services. It returns field performance, indexation, search visibility, and organic traffic data, with optional PDF/HTML reporting. Adoption requires local Python execution, network access, and appropriate Google credentials for most commands.
The skill first checks credentials with google_auth.py and identifies the available tier based on an API key, OAuth or service account access, GA4 configuration, and Google Ads credentials. It can then run PageSpeed Lighthouse and CrUX checks, 25-week CrUX history, Search Console analytics and URL Inspection, sitemap status checks, Indexing API submissions, GA4 organic traffic reports, YouTube searches, Google NLP entity analysis, Google Ads keyword ideas and volume lookups, Knowledge Graph checks, and Web Risk checks. Results can be emitted as structured JSON and passed to google_report.py for Core Web Vitals, Search Console, indexation, or full PDF reports.
- An SEO consultant needs real-user Core Web Vitals for a page rather than relying only on Lighthouse lab estimates.
- A site owner needs to verify Google index status, canonical selection, crawl state, or mobile usability for a URL.
- An SEO team wants the last 28 days of clicks, impressions, CTR, position, and high-impression quick-win queries from Search Console.
- A growth team wants daily organic sessions, users, pageviews, bounce rate, and engagement summarized from GA4.
- A technical SEO team needs batch URL inspections or URL submissions while tracking the relevant Google API quotas.
What are this skill's strengths and limitations?
- Uses Google’s own data sources for CrUX field metrics, Search Console performance, actual indexation status, and GA4 organic traffic.
- Its tiered credential model lets users start with API-key services and add Search Console, GA4, or Google Ads capabilities later.
- Includes 25-week CrUX trend analysis, batch URL inspection, structured JSON output, and PDF report generation.
- Released under the MIT license as part of an open-source claude-seo repository.
- Requires local Python scripts, a Google Cloud project, and appropriate permissions; many commands are unavailable without credentials.
- The Indexing API is officially intended for JobPosting and BroadcastEvent/VideoObject pages and has a 200-publish-request daily quota.
- Search Analytics data has an approximately two-to-three-day delay, and URL Inspection is limited to 2,000 requests per site per day.
- A CrUX 404 can mean insufficient Chrome traffic; without field data, PageSpeed lab data is the fallback.
- The SKILL.md does not establish direct use outside Claude Code; Codex or model-API deployments would need command, script, and runtime adaptation.
How do you install this skill?
The skill is installed as part of the full AgriciDaniel/claude-seo collection. For Claude Code 1.0.33+, run /plugin marketplace add AgriciDaniel/claude-seo, then /plugin install claude-seo@agricidaniel-claude-seo. Alternatively, run git clone --depth 1 https://github.com/AgriciDaniel/claude-seo.git followed by bash claude-seo/install.sh. The source does not document a standalone installation procedure for only skills/seo-google/SKILL.md.
How do you use this skill?
Configure Google credentials, then check them with python3 scripts/google_auth.py --check --json as documented by the skill. Example commands include /seo google setup, /seo google pagespeed <url>, /seo google crux <url>, /seo google crux-history <url>, /seo google gsc <property>, /seo google inspect <url>, /seo google sitemaps <property>, /seo google ga4 [property-id], /seo google youtube <query>, and /seo google nlp <url-or-text>. After collecting data, use /seo google report <type> to generate a report. Each command depends on the credential tier required by its API.
How does this skill compare with similar options?
This is not a complete crawler or full-site SEO audit tool. It supplies Google field data, indexation status, search performance, and analytics to complement the repository’s other SEO skills. Compared with Lighthouse-only analysis, it adds CrUX real-user measurements; compared with a manual Search Console workflow, it adds command-based querying, batch inspection, and report generation.