Longevity Protocol
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill follows health coaching best practices by prioritizing evidence-based pillars (sleep, exercise, diet) and explicitly labeling recommendations by scientific strength tiers. It contains no commands for file system access, network requests, or code execution. The instructions include robust escalation procedures for medical conditions and symptoms, ensuring clinical matters are referred to physicians.- [INDIRECT_PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface due to its requirement to process and analyze user-supplied health data.
- Ingestion points: The skill collects detailed personal data in Step 1 of SKILL.md, including age, sex, exercise habits, sleep patterns, dietary info, smoking/alcohol use, and current medications.
- Boundary markers: There are no explicit delimiters or 'ignore' instructions provided to the agent to protect against malicious instructions hidden within the user's health data inputs.
- Capability inventory: The skill is restricted to natural language generation and referencing other internal agent skills for specific planning; it has no access to subprocesses, shell commands, or network tools.
- Sanitization: No input validation or sanitization logic is present in the instructions.
Audit Metadata