Claude SEO Site Analysis
Run parallel, testable audits of a website’s technical, content, and AI-search performance.
The evidence shows conditional credential checks, error reporting, and README guidance on credential storage permissions and installer review. However, data flows, permission scope, costs, user confirmation, and rollback are not fully documented for crawling, APIs, report generation, baseline writes, or other external effects; the mandatory community footer also adds external promotion. No explicit malware, credential theft, or destructive-default red line is shown.
The evidence includes command documentation, conditional branches, and partial handling for unreachable URLs, failed sub-skills, and ambiguous business types. However, the supplied files cannot reproduce all referenced sub-skills, scripts, MCP services, or credentials; orchestration and command counts contain inconsistencies such as 24 versus 25, and abnormal-input diagnostics are limited. Static calibration therefore caps this below 10.
The evidence clearly describes target users, industries, commands, trigger terms, and many SEO scenarios. However, the scope is exceptionally broad, non-fit boundaries, input/output contracts, and anti-trigger conditions are incomplete, Chinese-language support is not stated, and core features may depend on Google, DataForSEO, Firecrawl, or other external services that may be difficult to reach from mainland China.
The evidence provides front matter, an MIT license, version metadata, quick reference material, progressive reference loading, installation notes, command tables, error handling, and some limitations. Maintenance ownership, update path, changelog, dependency versions, and parameter stability are incomplete, and documentation contains inconsistencies such as 24 versus 25 sub-skills, so points are deducted.
The evidence presents many concrete SEO tasks, structured reports, prioritized action plans, falsifiability checks, and multiple output formats; the README also includes a sample report and outcome claims. Nevertheless, the supplied material lacks reproducible evidence covering key execution paths, and actual output correctness, completeness, and cost-benefit remain unverified. Static calibration therefore limits this to a low score.
The evidence names several Google, Schema.org, DataForSEO, and OpenStreetMap sources and includes some formulas, methodology, and sample output. However, many detailed claims rely on author assertions or loosely identified studies, with no complete claim-level attribution, committed test coverage, or independent corroboration in the supplied material. Static calibration caps this below 5.
- This is a source-only static review; no commands, scripts, crawls, or API calls were executed. Review installer scripts, MCP permissions, external requests, and paid APIs before installation.
- Audits may write local reports, drift baselines, or caches and may incur third-party API charges; confirmation, cost ceilings, and rollback are not sufficiently specified.
- Reachability, quotas, credentials, and freshness of Google, DataForSEO, Firecrawl, and related dependencies must be verified in the target environment.
- The documentation contains inconsistencies in sub-skill counts and capability descriptions; validate command routing and output formats on key paths first.
What does this skill do, and when should you use it?
This is the `skills/seo/SKILL.md` skill from AgriciDaniel/claude-seo, designed for SaaS, local-service, e-commerce, publisher, and agency websites. It exposes `/seo` commands for full audits, page analysis, technical SEO, structured data, content quality, Core Web Vitals, sitemaps, and GEO. A full audit coordinates 24 sub-skills and up to 18 specialist agents, producing a 0–100 SEO Health Score, prioritized findings, and an action plan with validation checks. The core workflow can run without API credentials, while Google APIs, MCP extensions, and browser rendering add enrichment.
It reads a target website URL, detects the likely business type, and analyzes crawlability, indexability, INP-based Core Web Vitals, Schema.org markup, E-E-A-T, images, sitemaps, AI-search citability, local SEO, e-commerce SEO, backlinks, semantic clustering, and SEO drift. /seo audit dispatches technical, content, schema, sitemap, performance, visual, GEO, SXO, and conditional specialist agents, then produces Markdown reports, a prioritized action plan, and an optional PDF report. Google APIs, Firecrawl, DataForSEO, and similar integrations are used only when the relevant credentials or extensions are available.
- An SEO agency runs repeatable weekly audits across multiple client sites and monitors baseline changes.
- An in-house SEO lead at a SaaS, publishing, or e-commerce company checks technical issues, content quality, and AI-search readiness before a quarterly review.
- A freelance SEO consultant generates a health score and priority findings during client discovery or before writing a proposal.
- A local-service business audits Google Business Profile signals, NAP consistency, reviews, local schema, and map-related factors.
- A content team creates content briefs, semantic clusters, and internal-link recommendations around target topics.
What are this skill's strengths and limitations?
- Covers technical SEO, content quality, structured data, performance, GEO, local SEO, and e-commerce SEO.
- Full audits use parallel specialist agents and consolidate results into a health score and prioritized action plan.
- Recommendations include their observation basis, dependencies, failure checks, and monitorable leading indicators.
- The core workflow needs no paid API or per-domain subscription, is MIT-licensed, and can run locally.
- Supports Playwright rendering for most SPAs and optional Google API, Firecrawl, and other MCP integrations.
- The skill relies on Claude Code slash commands and subagent orchestration, so other clients require adaptation.
- Without Google APIs or external extensions, Core Web Vitals are primarily lab estimates and indexation is inferred from page signals.
- Scroll-triggered hydration, interaction-dependent content, and third-party widget timing can produce noisy SPA findings.
- DataForSEO, Ahrefs, Profound, SE Ranking, and similar enrichment require separate accounts or credentials.
- The supplied material gives repository-level test claims but does not provide independent test-coverage evidence for this individual skill.
How do you install this skill?
The public repository is MIT-licensed. For the Claude Code plugin installation, run /plugin marketplace add AgriciDaniel/claude-seo, then /plugin install claude-seo@agricidaniel-claude-seo. On Unix/macOS/Linux, you can instead run git clone --depth 1 https://github.com/AgriciDaniel/claude-seo.git followed by bash claude-seo/install.sh. On Windows, run powershell -ExecutionPolicy Bypass -File claude-seo\install.ps1. These commands install the repository’s plugin collection, after which the seo skill is available.
How do you use this skill?
In Claude Code, run /seo audit https://example.com for a full audit. Other supported examples include /seo page https://example.com/about, /seo technical https://example.com, /seo schema https://example.com, /seo content https://example.com, /seo geo https://example.com, and /seo sitemap generate. The repository documents the complete command and extension setup; without API credentials, Google live data and some external enrichment are unavailable.
How does this skill compare with similar options?
Compared with Screaming Frog, it emphasizes conversational LLM synthesis, recommendation validation, and AI-search analysis rather than deeper, faster link-graph crawling. Compared with Ahrefs Site Audit, it has no proprietary backlink index and can connect to Ahrefs through an extension; its differentiators are MIT licensing, no per-domain pricing, and local-first operation.