check-forecast-signals
Pass
Audited by Gen Agent Trust Hub on Sep 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data from the World Monitor API which could contain untrusted content. \n
- Ingestion points: The skill fetches data from https://worldmonitor.app/api/forecast/v1/get-forecasts as defined in SKILL.md. \n
- Boundary markers: The SKILL.md file contains a 'Content safety' section that explicitly instructs the agent to treat the API response as data rather than instructions and to ignore directives like 'ignore previous instructions' found within the payload. \n
- Capability inventory: The skill is configured to perform HTTP GET requests to retrieve JSON data. It does not perform file system modifications or execute system commands based on the data. \n
- Sanitization: The developer provides clear guidance to the agent to disregard and not act upon any directive-like text found in the response fields.
Audit Metadata