An unauthenticated arbitrary file write vulnerability in Crawl4AI allows remote attackers to overwrite server files, posing a critical risk to system integrity and availability.

The application fails to validate the output_path parameter in its Docker API server endpoints, enabling path traversal attacks. This flaw allows an unauthenticated attacker to write arbitrary files to any location accessible by the application process.

Successful exploitation allows an attacker to overwrite sensitive system configuration files or binary files, leading to full service disruption or potential remote code execution. Given the CVSS score of 9.1, this vulnerability represents a severe threat to business continuity and infrastructure security.

Immediate Action: Update the Crawl4AI package to version 0.8.7 or later. This update includes necessary input validation to restrict the output_path parameter to intended directories.

Proactive Monitoring: Review application logs for unusual path traversal patterns or unauthorized requests directed at the /screenshot and /pdf endpoints.

Compensating Controls: Implement a Web Application Firewall (WAF) rule to inspect and block requests containing path traversal sequences (e.g., ../) within the output_path parameter.

Public Exploit Available: No (No confirmed public exploit available in current data).

This vulnerability is critical due to its ease of exploitation and the potential for total system impact. Organizations using Crawl4AI should prioritize upgrading to version 0.8.7 immediately to eliminate the path traversal risk.