GPU VulnDB

Database/AI/ML frameworks & serving

Crawl4AI Docker API: unbounded delay fields let an unauthenticated client exhaust the crawler page pool

CVSS 8.7CVE-2026-119119AI/ML frameworks & servingcurated

Impact

_clamp_untrusted() fails to cap delay_before_return_html, mean_delay and max_range in untrusted CrawlerRunConfig bodies. A client posting huge values to /crawl or /crawl/stream holds browser pages open or stalls the rate limiter until the page pool is exhausted, after which every other crawl on that instance blocks. On a shared node this is a tenant-reachable way to park the service's browser workers indefinitely; recovery means restarting the container, and queued jobs are lost.

Who can reach it

Unauthenticated HTTP request to the Crawl4AI Docker API's /crawl or /crawl/stream endpoint. Anyone who can reach the API port.

What to do

Upgrade Crawl4AI to 0.9.5 or later and restart the API container. Interim mitigation is to front the API with an authenticated gateway that rejects or clamps these config fields and to rate-limit callers. Container image replacement and restart; no node reboot.

References

Related entries

All AI/ML frameworks & serving entries

This entry is curated: imported from vendor advisories with machine assistance, not yet individually verified. Confirm against your vendor's advisory before acting, and report anything wrong.