santa-method
Pass
Audited by Gen Agent Trust Hub on Apr 14, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill architecture creates a surface for indirect prompt injection within the verification workflow.
- Ingestion points: Untrusted data enters the agent context via the
{output}variable inside theREVIEWER_PROMPTtemplate defined inSKILL.md. - Boundary markers: Absent. The instructions do not use delimiters or explicit 'ignore embedded instructions' warnings for the interpolated output content.
- Capability inventory: The skill leverages the
Agenttool to spawn sub-agents (Reviewer B and Reviewer C) which then process the potentially malicious input. - Sanitization: Absent. The implementation documentation does not describe any validation, escaping, or filtering of the
{output}content before it is processed by the review agents.
Audit Metadata