Technical SEO audit
Crawl a site and get evidence-backed findings across seven categories: technical, indexability, structured data, international (hreflang), internal links, images and on-page. Findings are grouped by root cause.
Six capabilities built around one agency workflow: audit a client site, prioritise what to fix, monitor it between check-ins and report the work with evidence.
Crawl a site and get evidence-backed findings across seven categories: technical, indexability, structured data, international (hreflang), internal links, images and on-page. Findings are grouped by root cause.
Daily, weekly or monthly scheduled audits. Every completed audit is compared with the previous one and notifications flag new errors and URLs that started failing.
AI-search readiness checks that run with every audit: AI crawler access in robots.txt, content structure, entity consistency and source signals. No score, no promises.
Tracked questions, competitors and manually recorded AI answers, analysed for brand mentions, competitor mentions and website citations with exact excerpts as evidence.
Branded client reports built from stored audits: executive summary, root causes, broken-link priorities, GEO readiness, comparisons and an action plan. PDF, HTML and CSV.
One workspace per agency with Owner, Admin and Viewer roles enforced on the server, invitation links, data export and plan usage limits.
Each finding lists the facts it is based on: the status code, the rule line in robots.txt, the parser error, the character count. You can check it yourself.
There is deliberately no “health score”, no GEO score and no AI visibility score. Counts are shown with what they are counts of.
If a crawl didn’t reach a page, its issues are “not verified” or “not re-checked”, never silently “fixed”. Partial crawls are flagged.
Analysis and report wording are rule-based: the same data always produces the same report. No AI writes your client deliverables.
| Area | Current behaviour |
|---|---|
| Crawl size | Up to 1,000 URLs per audit in the dashboard. Proposed plans set 250 / 500 / 1,000 pages per audit. |
| JavaScript | Not rendered. The crawler analyses the HTML returned by the server; likely JavaScript shells are flagged. |
| Crawler etiquette | Respects robots.txt, identifies itself as SEOSentinelBot, waits between requests (0.5 s by default) and checks one site at a time. |
| Monitoring | Scheduled (daily, weekly, monthly), not real-time. Notifications appear in the dashboard; no email alerts yet. |
| AI platforms | Never contacted. AI visibility data is recorded manually; synthetic demo data is labelled and kept separate. |
| Structured data | JSON-LD is validated against a documented subset of Schema.org and Google rules; Microdata and RDFa are detected, not validated. |
| Language | Generic-anchor and content-readiness heuristics are English-oriented. |
The interactive demo runs the actual SEO Sentinel interface on a fictional agency and fictional client websites. No sign-up, nothing to install.
SEO Sentinel is in private beta. Nothing is for sale yet and no card is ever requested.