hunt-auth-bypass
Pass
Audited by Gen Agent Trust Hub on Nov 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill requires the agent to ingest and analyze untrusted content from external target applications, including SAML assertions, JavaScript source code, and HTTP response headers. This methodology exposes the agent to indirect prompt injection if the external data contains malicious instructions designed to subvert its logic.
- Ingestion points: Mapping authentication entry points, spidering targets, and capturing/decoding SAML responses (SKILL.md).
- Boundary markers: The instructions lack delimiters or specific warnings to ignore instructions embedded in the external data.
- Capability inventory: The skill uses network-capable tools like curl and subfinder, and includes Python snippets for processing the captured data.
- Sanitization: No instructions are provided for sanitizing or escaping content retrieved from target systems.
- [COMMAND_EXECUTION]: The skill contains multiple shell command templates (using curl and subfinder) and Python scripts intended for vulnerability detection and exploitation testing.
- Evidence: Templates for XMLRPC probing, SAML signature stripping, and JWT manipulation are provided as actionable methodology steps for the agent or user.
Audit Metadata