Qwen-Agent through 0.0.34 contains a server-side request forgery vulnerability in the document parsing path that treats caller-supplied paths as URLs without scheme restriction or host validation. Attackers can reach the unauthenticated Gradio interface to make the server issue HTTP requests to arbitrary internal addresses including metadata services and read retrieved content through parsed document output.

Qwen-Agent versions through 0.0.34 contain a server-side request forgery (SSRF) vulnerability in the document parsing component. The vulnerability arises because caller-supplied paths are treated as URLs without any scheme restriction or host validation. Attackers can exploit the unauthenticated Gradio interface to cause the server to issue HTTP requests to arbitrary internal addresses, including cloud metadata services. Retrieved content can be read back through the parsed document output, enabling internal network reconnaissance and sensitive data exfiltration. The vulnerable code resides in the simple_doc_parser.py file. This poses a significant risk in cloud-hosted deployments where metadata services expose credentials and configuration. No authentication is required to trigger the vulnerability.