hunt-exceptional-conditions
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest and analyze response bodies from external web endpoints to detect internal information leaks. This represents a potential tool output poisoning vector where an untrusted source (the targeted server) could return malicious instructions disguised as stack traces or error messages.
- Ingestion points: HTTP response bodies returned during endpoint fuzzing and testing (referenced in SKILL.md).
- Boundary markers: No specific delimiters or "ignore embedded instructions" warnings are provided for the agent when processing these external responses.
- Capability inventory: The agent is expected to perform network operations to send malformed data and may use file system capabilities to record findings as evidence.
- Sanitization: The instructions do not include steps to sanitize or validate the content of the response bodies before analysis.
Audit Metadata