How healthy is your website?
Audit any URL - pre-launch or live - for technical, SEO, and accessibility issues. Server-side static analysis driven from Elm, no headless browser.
What it checks
Rendering architecture (SSG / bot-aware SSR / CSR), title and meta tags, canonical, favicon, viewport, html lang, h1 and heading hierarchy, image alt text, HTTPS and HSTS, security headers including COOP, CSP quality, cookie security, server-version disclosure, perceived load time, render-blocking resources, compression, DOM size, static asset cache lifetimes, robots.txt content validity, sitemap, crawl directives, llms.txt for AI agents, form labels, landmarks, Open Graph and Twitter cards, JSON-LD, PWA metadata, internal and external link health.
Run from CLI, Claude, or any agent
POST https://webhealth.co/_r/audit with a JSON body {"url":"<your-site>"}. Poll the same endpoint until status is "ready", then read report.
curl -X POST https://webhealth.co/_r/audit \
-H "Content-Type: application/json" \
-d '{"url":"https://your-site.com"}'
JSON response. No auth, no API key.
Install as an agent skill (Claude Code, Codex)
# Claude Code
mkdir -p ~/.claude/skills/webhealth && curl -s -d '{}' \
https://webhealth.co/_r/skill -o ~/.claude/skills/webhealth/SKILL.md
# Codex CLI
mkdir -p ~/.codex/prompts && curl -s -d '{}' \
https://webhealth.co/_r/skill -o ~/.codex/prompts/webhealth.md
Then say /webhealth from any project: the skill infers the production URL from the repo, audits it, and can run the fix-deploy-re-audit loop with your approval.