Remote XML Schema Processing Arbitrary File Read in Guardrails-Detectors

Executive Summary

This vulnerability exists in the guardrails-detectors component, specifically in its file_type content detector. It allows a remote attacker to send an arbitrary XML Schema Definition (XSD) string to the service. The service processes this XSD without proper restrictions or sanitization.

Because the xmlschema.XMLSchema() function used to process the XSD resolves schema references over any URL scheme (like HTTP(S) or file://) by default, an attacker can craft a malicious XSD that causes the server to make requests to arbitrary URLs or read local files.

This can lead to server-side requests to attacker-controlled or sensitive internal resources, potentially exposing sensitive information such as cloud provider credentials or internal network services.

Useful 0 Not Useful 0

Impact Analysis

The vulnerability can impact you by allowing an attacker to cause the vulnerable server to perform unauthorized server-side requests or local file reads.

This can result in the disclosure of sensitive information, including cloud provider credentials or access to internal network services that should not be exposed.

Although the attacker does not receive direct responses (the exploitation is blind), the server-side actions can compromise the confidentiality of sensitive data and internal infrastructure.

Useful 0 Not Useful 0

Detection Guidance

This vulnerability can be detected by monitoring for requests to the guardrails-detectors service that contain XML Schema Definition (XSD) strings with external schemaLocation references. Specifically, look for requests that include <xs:import>, <xs:include>, or <xs:redefine> elements pointing to external URLs or local file paths.

Since the exploitation involves sending malicious XSD data to the detector service, network traffic inspection tools can be used to identify suspicious payloads containing such schema references.

Example commands to detect suspicious requests might include using network capture tools like tcpdump or tshark to filter HTTP POST requests to the detector service and then searching for XSD schemaLocation patterns.

• tcpdump -i <interface> -A -s 0 'tcp port <detector_port> and (((ip[2:2] - ((ip[0]&0xf)<<2)) - ((tcp[12]&0xf0)>>2)) != 0)' | grep -i 'schemaLocation'
• tshark -i <interface> -Y 'http.request.method == "POST" && http contains "schemaLocation"' -T fields -e http.file_data

Additionally, reviewing application logs for requests containing XSD payloads with external references can help detect exploitation attempts.

Useful 0 Not Useful 0

Mitigation Strategies

Immediate mitigation steps include restricting or sanitizing the XML Schema Definition (XSD) inputs processed by the guardrails-detectors component to prevent external schema references from being resolved.

Specifically, configure the xmlschema.XMLSchema() function to restrict external references by setting the allow parameter to 'sandbox' or 'local', which limits schema resolution to safe locations.

If configuration changes are not immediately possible, consider implementing network-level controls to block outbound HTTP(S) and file:// requests initiated by the detector service to untrusted locations.

Also, monitor and restrict access to the detector service to trusted users and networks only, reducing the attack surface.

Finally, apply any available patches or updates from the vendor or project that address this vulnerability as soon as they become available.

Useful 0 Not Useful 0

Compliance Impact

This vulnerability can lead to the disclosure of sensitive information such as cloud provider credentials or access to internal network services due to server-side requests or local file reads triggered by malicious XML Schema Definitions.

Such unauthorized disclosure of sensitive data can negatively impact compliance with common standards and regulations like GDPR and HIPAA, which require protection of sensitive information and prevention of unauthorized access.

Useful 0 Not Useful 0