# seo-programmatic

> Programmatic SEO planning and analysis for pages generated at scale from data sources, covering template engines, URL patterns, internal linking automation, thin content safeguards, and index bloat prevention with Google's 2025 Scaled Content Abuse enforcement context.

**Use case**: Build programmatic pages at scale without triggering penalties

**Canonical URL**: https://agentcookbooks.com/skills/seo-programmatic/

**Topics**: claude-code, skills, marketing, seo

**Trigger phrases**: "programmatic SEO", "pages at scale", "dynamic pages", "template pages", "generated pages"

**Source**: [AgriciDaniel](https://github.com/AgriciDaniel/claude-seo/tree/main/skills/seo-programmatic)

**License**: MIT

---

## What it does

`seo-programmatic` is a Claude Code skill from AgriciDaniel's [claude-seo repo](https://github.com/AgriciDaniel/claude-seo). It analyzes and plans SEO pages generated at scale from structured data — integration pages, glossary pages, location + service combinations, product directories, and template-driven content. The skill enforces quality gates that reflect Google's 2025–2026 Scaled Content Abuse enforcement: a warning at 100+ pages without content review, a hard stop at 500+ pages requiring explicit justification, a 40% unique content minimum threshold (30% triggers a hard stop risk assessment), and a standalone value test for every template.

Beyond quality gates the skill covers: data source assessment (CSV/JSON/API/database record quality, duplicate detection), template design review (variable injection points, static vs dynamic content blocks, mad-libs detection), URL pattern strategy, internal linking automation (hub/spoke model with BreadcrumbList schema), canonical strategy for parameter variations, sitemap integration with `lastmod` accuracy requirements, and index bloat prevention (pagination noindex, faceted navigation canonicalization, crawl budget for 10k+ page sites).

## When to use it

Reach for it when:

- You are building a city + service page matrix and want to know how to structure templates that pass the quality gates
- You inherited a programmatic site and need to assess thin content risk and index bloat before doing any optimization
- You want to plan the rollout cadence for a large batch of generated pages (the skill recommends 50–100 page batches with 2–4 week monitoring intervals)

When *not* to reach for it:

- You are building fewer than 20 pages — the quality gate thresholds and batch rollout guidance is designed for genuine scale
- The content is fully unique, human-written articles — this skill is specifically for template-driven generation patterns

## Install

Copy the [`seo-programmatic` SKILL.md](https://github.com/AgriciDaniel/claude-seo/tree/main/skills/seo-programmatic) into `.claude/skills/seo-programmatic/`.

Trigger phrases: "programmatic SEO", "pages at scale", "dynamic pages", "template pages", "generated pages", "data-driven SEO".

Invoke with `/seo programmatic <url>` to audit an existing programmatic site, or `/seo programmatic plan` to plan a new one.

## What a session looks like

A typical session has three phases:

1. **Data source and template assessment.** The data powering the pages is evaluated: record count, column uniqueness, missing values, duplicate or near-duplicate records (above 80% field overlap is flagged). The template is reviewed for mad-libs patterns (identical text with only city/keyword names swapped) and conditional logic that adds genuine per-record value.
2. **Quality gate evaluation.** The skill calculates unique content percentage (words unique to this page / total words, excluding shared headers/footers/navigation). Pages below 40% unique content are flagged. Pages with less than 30% unique content trigger a hard stop. The scale thresholds from Google's June/August 2025 enforcement waves are cited with the 45% reduction in low-quality content Google reported.
3. **Architecture and rollout plan.** URL pattern strategy is designed. Internal linking automation is specified (hub pages, 3–5 related items per page, BreadcrumbList schema). Canonical strategy handles pagination and parameter variants. Progressive rollout is specified: publish in 50–100 page batches, monitor indexing and rankings for 2–4 weeks before expanding.

## Receipts

**Works well:** The mad-libs detection prompt — "if you can swap the city name and the content still makes sense, it is a doorway page" — is a clear and actionable test that most site builders can apply themselves without a tool. The skill gives you this test directly as part of the template review.

**Backfires:** The quality gate thresholds are conservative by design. A site with 35% unique content per page may still rank well in practice depending on the competitive landscape. The skill enforces the conservative threshold and requires acknowledgment — it does not adjust for competitive context.

**Pattern that works:** Build the hub pages and the template simultaneously, not separately. Hub pages need to know the template structure to link correctly, and templates need to know the hub structure to link back. Building them in sequence creates rework.

## Source and attribution

Originally written by [AgriciDaniel](https://github.com/AgriciDaniel). The canonical SKILL.md and supporting files live in the [`seo-programmatic` folder](https://github.com/AgriciDaniel/claude-seo/tree/main/skills/seo-programmatic) of the [claude-seo repository](https://github.com/AgriciDaniel/claude-seo).

License: MIT. Install, adapt, and redistribute with attribution preserved.

This page documents the skill from a practitioner's perspective. For the formal spec and updates, defer to the source repo.