fda-database
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill operates entirely as described, serving as a programmatic interface for medical regulatory research.
- [DATA_EXFILTRATION]: Network communication is limited to the official, well-known domain
api.fda.govfor its intended primary purpose. - [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources (FDA API). While this constitutes an ingestion surface, the data retrieved consists of structured regulatory records from a trusted government entity, representing minimal risk to the agent context.
- [DYNAMIC_EXECUTION]: The skill uses local file system caching to improve performance. Responses are stored as JSON files using deterministic MD5 hashes for filenames and are processed using standard safe JSON parsing libraries.
Audit Metadata