longevity
Pass
Audited by Gen Agent Trust Hub on Apr 8, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill facilitates indirect prompt injection by instructing the agent to ingest and analyze untrusted external data.
- Ingestion points: The skill explicitly directs the agent to process user-uploaded files (lab results, research papers) and external web content from PubMed and other sources in its 'Extended Capabilities' section.
- Boundary markers: There are no instructions provided to use delimiters or ignore embedded commands when processing these external data sources, which could lead the agent to follow malicious instructions hidden within research text or metadata.
- Capability inventory: The skill encourages the use of 'Web search' and 'File reading' tools to fulfill its primary research purpose, which provides the necessary channel for untrusted data ingestion.
- Sanitization: The instructions do not define any sanitization, validation, or filtering logic to mitigate the risk of processing adversarial content within the evidence being evaluated.
Audit Metadata