/seo-page

name: seo-page
description: Single-URL SEO review — runs the technical audit and the content audit on one page (on-page elements, meta, E-E-A-T signals, readability, keyword placement, schema, images, lab performance risks, internal links) in one pass with two deterministic scores (real Core Web Vitals field data when seo-google is connected). Lightweight alternative to seo-audit when only one page matters. Trigger when the user says "analyze this page", "check page SEO", "single URL", "check this page", "review this URL", or provides one specific URL.

seo-page

**Family:** seo **Status:** Stable

Purpose

The single-page analyzer: the same on-page, meta, schema, image, and internal-link checks as a full audit, scoped to one URL, plus observable performance issues — real Core Web Vitals *field* data only when `seo-google` is connected. Cheaper to run and easier to act on for spot checks, pre-publish reviews, and "is this page good?" questions. The free path (fetch + parse + score) is the whole product; paid APIs only add live ranking and field-performance context.

Triggers

• "analyze this page" / "check this page" / "review this URL"
• "single URL SEO" / "check page SEO"
• "page-level analysis"

Inputs

• A single URL (or local HTML file)
• Optional: target keyword(s) for relevance scoring

Steps

1. **Fetch the page** (WebFetch, or `curl`-style retrieval of the raw HTML). For a local build, read the file directly. 2. **On-page, meta & technical** — run the technical audit on the saved HTML (pass `--url` for URL-structure checks):

   python3 "${CLAUDE_PLUGIN_ROOT}/scripts/seo/tech_audit.py" --file page.html --url <URL> --no-network --human

Title, meta description, one `<h1>`, canonical, meta robots / X-Robots-Tag, Open Graph, viewport, lab CWV risks, JS-rendering and mixed content — each with a fix (catalog: `references/seo-technical/check-catalog.md`). 3. **Content quality** — run the content audit with the target keyword:

   python3 "${CLAUDE_PLUGIN_ROOT}/scripts/seo/content_audit.py" --file page.html --url <URL> --keyword "<target>" --type <article|product|local|home|category> --human

E-E-A-T signals (YMYL-aware), heading order, readability, depth for the page type, keyword placement / stuffing, in-content internal links and anchors, scaled-content tells, passage citability (rubric: `references/seo-content/content-rubric.md`). Then judge intent match by hand; hand deep E-E-A-T work to `seo-content`. 4. **Schema** — detect JSON-LD; validate type appropriateness and required properties; flag rich-result eligibility. (Delegate deeper work to `seo-schema`.) 5. **Images** — alt text presence/quality, dimensions declared (CLS), modern formats (WebP/AVIF), and lazy-loading below the fold. 6. **Performance** — note render-blocking resources and obvious payload issues. If `seo-google` (PageSpeed/CrUX) is available, pull real LCP/CLS/INP; otherwise report observable issues and say field data needs `seo-google`. 7. **Internal links** — the content audit counts in-content internal links and flags generic anchors; judge whether they point at the right hub and next step. 8. **Optional enrichment** — if the DataForSEO extension is configured, add live ranking/keyword context; otherwise state, in one line, what it would add. 9. **Render a compact report** — findings grouped Critical / High / Medium / Low, each with a specific fix.

Outputs

• Page-level findings grouped by priority, each with a concrete fix
• Two deterministic scores — technical (lab) and content — plus the combined view
• A short "what paid data would add" note when running the free path

Dependencies

• `scripts/seo/tech_audit.py` (required) — technical / on-page / lab-CWV checks
• `scripts/seo/content_audit.py` (required) — content quality, E-E-A-T signals, links
• `references/seo-content/content-rubric.md` (required) — content thresholds + scoring
• Optional: `seo-schema` (deep schema), `seo-content` (deep E-E-A-T),

`seo-google` (real CWV/field data), DataForSEO extension (ranking context)

Notes

A focused single-page subset of `seo-audit`; both exist because their cost profiles differ (one URL vs. a full crawl). They share the same checks — keep the simple entry point (see `references/ENGINE-CONTRACTS.md` §5 on the free path).