santa-method
Pass
Audited by Gen Agent Trust Hub on Apr 11, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The 'Santa Method' framework introduces a risk of indirect prompt injection as it requires agents to evaluate untrusted content.\n
- Ingestion points: Untrusted content and task specifications enter the agent context via the
{output}and{task_spec}variables in theREVIEWER_PROMPT(SKILL.md).\n - Boundary markers: The prompt uses markdown headers in SKILL.md to separate the output under review, but fails to include instructions that explicitly command the agent to ignore directives within that content.\n
- Capability inventory: The skill utilizes the
Agenttool (documented in SKILL.md) to perform reviews and corrections; a successful injection could compromise these sub-processes.\n - Sanitization: No sanitization, escaping, or validation of external content is implemented in the described framework.
Audit Metadata