us-property-data

Pass

Audited by Gen Agent Trust Hub on Sep 1, 2026

Risk Level: SAFEDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [DATA_EXFILTRATION]: The skill performs outbound HTTPS requests to the external domain api.zillapi.com. It manages authentication by reading the ZILLAPI_KEY environment variable and transmitting it as a bearer token in request headers.
  • [INDIRECT_PROMPT_INJECTION]: The skill acts as a conduit for processing untrusted external data, specifically property identifiers provided by users.
  • Ingestion points: Address strings, Zillow URLs, and property IDs (zpids) provided by the user and processed in SKILL.md.
  • Boundary markers: The skill lacks technical delimiters (like XML tags) for input data, though it provides instructional guidance for the agent to verify estimates and handle missing data points.
  • Capability inventory: The skill allows the agent to execute GET and POST network requests to the api.zillapi.com domain.
  • Sanitization: The instructions explicitly direct the agent to minimize data transmission by sending only required identifiers and avoiding the inclusion of personal data or credentials in parameters.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 1, 2026, 07:28 PM
Security Audit — agent-trust-hub — us-property-data