amazon-analysis
Pass
Audited by Gen Agent Trust Hub on Aug 25, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes untrusted data from Amazon product listings and consumer reviews to generate market insights. This constitutes an indirect prompt injection surface, which is a standard risk for research-oriented skills.\n
- Ingestion points: External content is retrieved through the realtime/product and reviews-raw subcommands in scripts/zoodata.py.\n
- Boundary markers: The prompt templates in the Python script use delimiters but do not include explicit instructions for the agent to ignore instructions embedded within the user-generated review content.\n
- Capability inventory: Capabilities include network access to the vendor API and local file system read access for configuration.\n
- Sanitization: No specific escaping or filtering of external text was identified in the provided prompt logic.\n- [DATA_EXPOSURE_AND_EXFILTRATION]: The skill communicates with the official vendor API at api.zoodata.ai to retrieve data. The Python CLI enforces a trusted-host check to ensure the Bearer token (ZOODATA_API_KEY) is only transmitted to verified zoodata.ai hosts or localhost. The script also accesses the local configuration file at ~/.zoodata/config.json for credential management, which is the documented and intended behavior.
Audit Metadata