The vulnerability resides in the SSRF protection logic of the Crawl4AI Docker API server, specifically in the URL validation used to guard webhook callbacks. The filter relied on a hard‑coded IPv4 and IPv6 CIDR blocklist that attacker to encode an internal IPv4 address inside an IPv6 transition form or use the IPv6 unspecified address. This bypass allows the crawler to issue requests to internal services and cloud metadata endpoints such as 169.254.169.254, thus exposing sensitive data or internal network resources. The flaw is a form of input validation.

All installations of the open‑source Crawl4AI LLM‑friendly web crawler with Docker deployment.8 are affected. The product is provided by the vendor unclecode. Versions 0.8.0 through 0.8.7 do not contain the fix; any release 0.8.8 or newer includes the remediation.

The CVSS score of 7.5 indicates a high severity. Because the Docker API server is unauthenticated by default (jwt_enabled is false), no credentials are required; any host that can reach the API endpoint can trigger the SSRF. The exploit can be performed by sending a crafted URL to the server, which will then fetch internal resources, potentially leaking data or enabling further attacks. The EPSS score is not available, but the lack of authentication makes the vulnerability very likely to be exploited in environments where the API is exposed. The issue is not listed in the CISA KEV catalog.