# pSEO Desk

> Paste the page template behind your programmatic SEO pages, with {{column}} placeholders, and the data rows that fill it - integration, comparison, location, glossary or directory pages - plus the business, the goal and the data source. The browser renders every page and measures it: the share of each page that is template text, near-duplicate pages, duplicate URLs and titles, the same comparison in both orders, empty placeholders, thin pages, missing links and schema, subdomain and unslugged URLs, zero-volume rows and the data source, with 22 flags across six areas, free; then an audit says whether to launch, and a template lane writes a stronger template that the page renders over the same data.

URL: https://pseo-desk.skillsafe.ai/
API tutorial: https://pseo-desk.skillsafe.ai/api.html

## Who it is for

Founders, marketers and SEO leads about to publish many similar pages from one template and a data table - "[tool] integration", "[X] vs [Y]", "[service] in [city]", "what is [term]", "best [category]" - who want to know before launch whether each page carries its own value or the same text with one word swapped, and what to change.

## The free scan (in the browser, no model)

- A small template language: header lines URL:, Title:, Meta:, H1:, Schema:, then the body after ---; {{column}}, {{column|slug}}, |lower, |upper, |title; {{#column}}...{{/column}} shown when the row has a value, {{^column}}...{{/column}} when it is empty. A full HTML page is read from its title, meta description, h1 and canonical link.
- Data as CSV, TSV, semicolon or pipe separated, or a Markdown table, with quoted cells.
- Every page rendered: URL, title, meta description, H1, body word count, the share of body words that are template text, and the nearest other page by shared three-word phrases. A preview of any page as a reader gets it, and a CSV of every page.
- The playbook (one of the skill's 12) and the keyword pattern; the data source on the skill's defensibility hierarchy (proprietary, product-derived, user-generated, licensed, public); search volume per row from a volume column.
- Six areas: data, template, unique value, technical SEO, internal linking, keywords and indexation.
- Flags PS-01..PS-22: placeholders with no column, a template that does not parse, duplicate URLs, empty placeholders, near-duplicate pages, mostly-template pages, thin pages, no conditional content, a missing title, meta, H1 or URL pattern, shared titles, comparisons in both orders, long titles and odd-length metas, one meta on every page, no internal links, no hub or related links, no schema, subdomain URLs, unslugged URLs, public data, zero-volume rows, doorway-style location pages, unused data columns.

## The paid lanes (one system prompt, routed on `task`)

- `audit`: the plan (playbook, keyword pattern, a one-line read), the six areas (good, weak, bad or unknown, each with a finding, quotes and a fix), the rendered pages in the page's own numbers, issues as finding, impact and fix with the page's flag ids, an index plan (index first, hold back, sitemap), 3 to 6 fixes, and a verdict: launch, fix first or rethink.
- `template`: a stronger page template - a slugged subfolder URL, title, meta and H1 patterns, 4 to 8 sections with conditional content driven by the data, the unique value per page, up to 6 new columns to add, hub-and-spoke linking with breadcrumbs, schema, and an indexation rule. The page renders it over the same data and can load it back into the template box for a free re-scan. It can carry the audit's findings by a handoff button.

Every reply is checked against the input: quotes word for word, the rendered-page numbers equal to the page's own render, flags carried, every placeholder a real or declared column, and no invented count, price, volume or ranking.

## Source

Derived from the agent skill @coreyhaines31/programmatic-seo (https://skillsafe.ai/skill/@coreyhaines31/programmatic-seo), by Corey Haines, in coreyhaines31/marketingskills (MIT): https://github.com/coreyhaines31/marketingskills/tree/main/skills/programmatic-seo - its core principles, the 12 playbooks (references/playbooks.md), choosing a playbook, the implementation framework (keyword pattern research, data requirements, template design, hub-and-spoke internal linking, indexation strategy), the pre-launch quality checklist and the common mistakes.

Runs on gpt-terra. Part of SkillSafe (https://skillsafe.ai/).