Most SEO slips are boring ones: a page with no meta description, two H1s, images without alt text, a title that gets cut off in results. They are easy to fix and just as easy to forget when you are editing at the end of a long day.
I wrote a small Hermes skill, seo-page-audit, that checks those basics on a single URL or a local HTML file. It is one Python file using only the standard library, so there is nothing to install and no API key.
It reads the HTML, then lists issues ranked high, medium or low: missing title or description, no H1 or several, heading level jumps, canonical, lang, viewport, noindex, Open Graph tags, image alt and dimensions, links with no text, and a rough word count. The exit code is 1 when there is a high-priority issue, so you can drop it into a CI step.
python3 scripts/seo_page_audit.py https://example.com/
On that page it reports a missing meta description and H1 as high priority, plus a missing canonical and thin content as medium. That matches what the page really contains.
It has limits, and the skill says so: it reads the HTML as served (no JavaScript rendering), the length thresholds are rules of thumb, and it audits one page, not a whole site. Treat it as a fast first pass, not a verdict.
The skill includes a --selftest that spins up a local server and checks both a bad and a clean page.
Skill (MIT): https://github.com/RobinBeraud/hermes-skills/tree/main/seo/seo-page-audit
All skills: https://github.com/RobinBeraud/hermes-skills
This post was written with AI assistance and reviewed through automated checks.
Top comments (0)