Crawler logs show which AI agents actually requested pages on your site — distinct from two other things Stride measures. Robots.txt checks show what agents are allowed to access; scans show what AI answers say about you. Crawler logs show the visits themselves.
What it measures
Your CDN or edge worker sends Stride each request a known AI agent makes: which agent, which path, when, and whether it succeeded. Stride matches the agent's identity and separates two kinds of activity:
- Verified agents — visits whose source IP checks out against the operator's published identity (reverse DNS or IP ranges).
- Self-identified agents — visits whose user-agent claims to be a known agent but whose identity could not be checked. These are labeled as such and never counted as verified.
Fetches triggered when a user opens a cited link in ChatGPT, Claude, or Perplexity appear as their own category. They are a real-time sign that people are opening links to your site from AI answers — not proof that any monitored answer cited you.
What it does not do
- Crawler visits never change your Visibility Index, scores, or scan results.
- A crawl is not a citation and never implies causation.
- Stride stores request paths only — no query strings, no visitor personal data. Raw events are deleted after 30 days.
Connecting
- Open Dashboard → Settings → Integrations and expand AI crawler logs
- Create an ingest key and copy it — it is shown only once
- Follow the setup guide for your platform
Available on Scale and Agency plans during beta.
Platform support
Cloudflare (via a small Worker you install) and Vercel (via Log Drains) are supported. Platforms that expose neither raw logs nor edge workers — Shopify, WP Engine, Kinsta without Cloudflare in front — are not supported yet. Putting Cloudflare in front of the site is the usual workaround.
Vercel Log Drains are batched and can drop entries, so numbers from that source are labeled as sampled.