Static-Site SEO Regression Gate
A full-site audit flags hundreds of pre-existing issues and nobody reads it.
$29
Works with the AI Agents you already use
About this skill
Sites rebuilt by generators and AI writers degrade in small steps: a template drops the Spanish hreflang, a draft ships with a placeholder title, a link to a renamed page breaks, the sitemap's lastmod is bumped on pages that did not change, a noindex slips onto a money page. A full-site audit flags hundreds of pre-existing issues and nobody reads it. This gate is incremental.
How it judges (seo_gate.py): the candidate is an overlay on the live tree, so links from changed pages to unchanged pages resolve; only pages whose bytes changed are checked; for each, the old version's findings are computed too, and only NEW findings block. Findings the live site already had are reported separately as inherited and never excuse a new one. A plan file from your generator restricts the check to listed files and refuses a candidate whose SHA-256 changed after generation.
Page checks: title and description length, one H1, accepted language, self-canonical, placeholder title, heading skips, over-long paragraphs, invalid JSON-LD, missing internal scripts, styles and images, missing alt, javascript: and data: links, nofollow on your own domains, broken internal links and anchors, hreflang targets missing, non-canonical or non-reciprocal, robots changes on existing pages, removed hreflang, removed protected sections, new pages without their language pair, duplicate titles across the served tree.
Sitemap checks: wrong origin, missing page, non-indexable page listed, lastmod bumped without a byte change, future or invalid dates, duplicates, removed URLs, folder replacements that would delete a listed page, new legal sub-pages.
Tunable thresholds and lists (--languages, --single-language, --service-prefixes, --owned-domains, --protected-ids, --legal-prefixes, --max-paragraph-words, title and description ranges). JSON output, --report with inherited findings. 20 unit tests, each exercising a real regression. Python 3.10+, standard library only, no network.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 17 days ago
Requires
Permissions
Allowed Hosts