Threat Intelligence Database

Crawl4AI versions before 0.8.8 have credential exfiltration vulnerabilities in their Docker API server. Attackers can exploit unauthenticated endpoints (/md, /llm, /llm/job) by manipulating the base_url parameter and setting api_token to environment variable references, allowing them to read arbitrary environment variables. This can lead to exfiltration of sensitive data such as provider API keys and JWT SECRET_KEY, enabling authentication bypass.

Crawl4AI versions before 0.8.7 have an arbitrary file write vulnerability in the Docker API server's /screenshot and /pdf endpoints. The vulnerability arises because the output_path parameter does not validate input, allowing attackers to write files to arbitrary locations accessible by the application user. This can lead to overwriting server files and cause denial of service.

Crawl4AI before 0.8.7 contains a server-side request forgery (SSRF) vulnerability in the Docker API server's /crawl/job and /llm/job endpoints, which accept webhook URLs without destination validation. An attacker can supply webhook URLs pointing to private or internal IP ranges, Docker networks, or cloud metadata endpoints (e.g. 169.254.169.254), causing the server to make requests to internal services and potentially expose cloud metadata.