Export logs through the Cloudflare Logpull API

[jan-dh]

With the rise of AI and the impact on SEO there's an emerging strategy: Server-side Query Fan-Out Tracking.
The idea is to look at your Cloudflare log files to gain insights on when your website was visited by AI bots (in a search) and to gain some insights on how strong your content is for certain topics etc. Interesting read in the details: https://conversem.com/the-query-fan-out-session/

There's already a script available that does a lot of the analysis but it needs the Cloudflare logs; the script can fetch them through the Cloudflare Logpull API. I'm not sure if this would be possible in Craft Cloud?

[timkelty]

Ohh interesting idea.

We plan on providing traffic metrics soon, so this could be a good companion. Putting it in the backlog!

[timkelty]

@jan-dh for your use-case, are you thinking core request details would be adequet (eg path, headres/user-agent) or are you thinking of the Cloudflare hueristic data (eg bot score, ja3 fingerprint)?

[jan-dh]

This is an example of the data we'd need:

{
    "EdgeStartTimestamp": "2025-12-13T00:00:36.625637+00:00",
    "ClientRequestURI": "/docs/latest/integrations",
    "ClientRequestHost": "docs.example.com",
    "ClientRequestUserAgent": "Mozilla/5.0 (compatible; Google-Extended; +https://developers.google.com/search/docs/crawling-indexing/google-common-crawlers)",
    "BotScore": 4,
    "BotScoreSrc": "Machine Learning",
    "VerifiedBot": true,
    "BotTags": [
        "llm"
    ],
    "ClientIP": "10.58.92.172",
    "ClientCountry": "US",
    "EdgeResponseStatus": 200
}

[timkelty]

@jan-dh if you need this before we officially support it: An easy workaround might be to use a (normally free) Cloudflare zone for your DNS, so you can ingest request logs before it proxies on to Craft Cloud.

[jan-dh]

Thanks, I'll pick this up with our ops team to see if this would be an option 🙏