SEO Content Audit Skill
Run a systematic SEO audit on any page or site and get an actionable report with prioritized fixes.
Read-only webpage analysis plus Markdown report generation; no credential access, no covert exfiltration, no destructive defaults; dependencies are stdlib plus optional reportlab, a small attack surface. Deductions: analyze_page.py is not included in the evidence, so its behavior (how it requests target sites) is not statically verifiable; robots compliance for crawled sites, confirmation mechanisms, and data-flow disclosure are unstated; publisher unverified, attribution incomplete.
The SKILL.md flow is self-consistent (script gathers data, then manual checklists, then templated output), and requirements.txt claims stdlib-only core scripts, reducing dependency risk. Deductions: the script itself is absent, so its error handling and failure feedback cannot be verified; no tests or key-path reproduction; handling of abnormal inputs (anti-bot walls, timeouts, non-HTML pages) is entirely unspecified, so no high static score.
Trigger conditions are clear (URL plus SEO-analysis request, or /market seo), output format is explicit, scenario definition is decent. Deductions: no declared capability boundaries (e.g., JS-rendered SPAs, login-gated pages are not suitable); content is English-only with no Chinese-language support; reachability of crawled pages from mainland-China networks is uncertain.
Well-layered docs: purpose, triggers, steps, output template, principles; MIT license; uninstall path exists. Deductions: no version number or changelog for this skill; no known-limitations disclosure or FAQ; the scripts/analyze_page.py path is only indirectly corroborated by the README; maintenance ownership and update path unclear (publisher unverified).
The output template is concrete (SEO-AUDIT.md with health score, prioritized recommendations, E-E-A-T table), directly usable in principle, and methodology follows public SEO best practices with clear marginal value. Deductions: not executed, so representative outputs unverified; several quantified benefit claims (CTR +20-35%, 24% abandonment reduction) lack sources and need review; the health score (X/100) rubric is undefined.
Audit criteria and benchmarks (LCP<2.5s etc.) broadly match public Google documentation and are partially corroborable. Deductions: no tests, no CI coverage evidence, no real execution samples; the core script is not in evidence so key paths are not statically reproducible; some research-style statistics are untraceable to sources.
- Static review; no code executed. analyze_page.py is not in the evidence, so its network behavior, error handling, and failure feedback are unverified.
- README claims like '5 parallel subagents' are repository-level descriptions; this score covers only the market-seo skill.
- Quantified SEO benefit claims (CTR uplift, CWV impact) lack sources; verify independently before relying on them.
- English-only content, and reachability of crawled pages from mainland China is uncertain; sites with restrictive robots.txt or anti-bot measures may fail.
- Publisher is unverified in the FollowSkills registry; maintenance ownership and update path unclear.
- No skill-level version or changelog; diff sources before upgrading.
What does this skill do, and when should you use it?
This is the SEO content audit skill from the ai-marketing-claude suite, triggered via /market seo <url>. It first runs a bundled Python script (analyze_page.py) to extract baseline page data — title, meta tags, heading hierarchy, images, links, schema — then walks through an 11-step expert review covering on-page checklists, E-E-A-T content quality scoring, keyword and search intent analysis, technical SEO, content gaps, featured snippets, Core Web Vitals, and content strategy. The output is an SEO-AUDIT.md with a health score out of 100 and recommendations tiered by priority. Install the full suite to use it; MIT licensed.
Runs python3 scripts/analyze_page.py <url> to extract structured page data (title tag, meta description, H1-H6 hierarchy, internal/external links, image alt status, schema, canonical, robots directives); then scores on-page elements (title, meta description, headings, images, internal linking, URL structure) on Pass/Needs Work/Fail checklists; rates content against Google's E-E-A-T dimensions (Experience, Expertise, Authoritativeness, Trustworthiness); analyzes primary keyword placement, 1-2% density benchmarks, and search intent alignment; quick-checks robots.txt, sitemap, canonical tags, page speed benchmarks (LCP/FID/CLS/TTFB/FCP), and mobile-friendliness; performs content gap analysis, featured snippet optimization, schema markup audit, and internal linking opportunity identification, plus a Core Web Vitals assessment with revenue-impact figures and content strategy recommendations; finally writes SEO-AUDIT.md with a health score and Critical/High/Medium/Low priority recommendations.
- An SEO consultant auditing a prospect's site before a pitch, producing a client-ready audit document
- A content site owner self-checking keyword placement, E-E-A-T signals, and intent match after publishing
- An agency establishing a baseline SEO health score for a new client and sequencing fixes by effort-to-impact
- An indie founder diagnosing technical SEO issues (robots, sitemap, canonical, Core Web Vitals) on their own landing page
- A content team finding content gaps and featured snippet opportunities to plan next quarter's editorial calendar
What are this skill's strengths and limitations?
- Comprehensive coverage: 11 steps spanning title tags to Core Web Vitals form a systematic methodology
- Combines automated data extraction with expert-level interpretation rather than raw script output
- Structured, deliverable output: SEO-AUDIT.md with health score, before/after examples, and four priority tiers
- Business-outcome framing — recommendations include quantified impact estimates (e.g., 20-35% CTR lift)
- MIT license allows free modification and commercial use
- Core Web Vitals section provides benchmark tables only; no actual speed-testing tool (e.g., PageSpeed API) is invoked
- Keyword search volume, competition, and People Also Ask data must be sourced externally; the skill doesn't fetch them
- No test suite or accuracy validation is documented in the source
- Audit depth depends on the model's ability to read the page; handling of JS-rendered or login-walled pages is unspecified
- One of 15 bundled skills — you install the suite even if you only want this one
How do you install this skill?
Install the whole suite (this skill is bundled): curl -fsSL https://raw.githubusercontent.com/zubair-trabzada/ai-marketing-claude/main/install.sh | bash, or manually git clone https://github.com/zubair-trabzada/ai-marketing-claude.git and run ./install.sh. The skill lives at skills/market-seo/SKILL.md with the analyzer at scripts/analyze_page.py. pip install reportlab is optional and only needed for the suite's PDF reports (this skill outputs Markdown). The installer places files under ~/.claude/skills/ and ~/.claude/agents/.
How do you use this skill?
In Claude Code, type /market seo <url> (or /market seo and provide the URL when prompted). The skill runs python3 scripts/analyze_page.py for page data, then works through the 11-step analysis and writes SEO-AUDIT.md to the working directory. If you've previously run /market audit or /market landing, findings are cross-referenced. The source does not document which Python parsing libraries are required or their versions.
How does this skill compare with similar options?
Within the same suite, /market audit runs a six-dimension overall marketing audit where SEO is only 20% of the weighted score — this skill is the deep, SEO-only alternative; /market quick is the fast-snapshot option.