The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill contains hardcoded references to a testing payload URL (http://attacker.example.com/xxe.dtd) in SKILL.md. While used for security verification, this domain is flagged as suspicious by automated reputation scanners.
[INDIRECT_PROMPT_INJECTION]: The skill fetches and parses external XML data (WSDL definitions) to configure its testing parameters, creating a potential vector for indirect prompt injection.
Ingestion points: WSDL file content is fetched from a user-provided URL in scripts/agent.py and SKILL.md.
Boundary markers: No delimiters or safety instructions are used to separate the external XML data from the tool's execution context.
Capability inventory: The skill uses requests to perform network operations and lxml to parse XML content.
Sanitization: The skill uses lxml.etree.fromstring() to parse remote XML content without explicit security configurations to disable DTD or entity resolution, which could expose the tool itself to XXE attacks if the source WSDL is malicious.
[COMMAND_EXECUTION]: The skill generates and sends exploit payloads (SQL injection, XXE, SOAPAction spoofing) to remote endpoints. These actions are the primary purpose of the skill and are performed against targets specified by the user.

performing-soap-web-service-security-testing