weather

Pass

Audited by Gen Agent Trust Hub on Aug 1, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill makes network requests to wttr.in, a well-known and widely used public weather service. These requests are limited to fetching weather data in JSON, text, or PNG formats and do not involve downloading or executing arbitrary code.
  • [COMMAND_EXECUTION]: The skill uses curl to interact with the weather service. These commands are static and do not incorporate unvalidated user input in a way that would allow for command injection. The Python script also uses standard library modules for its operations.
  • [DATA_EXFILTRATION]: There is no evidence of sensitive data access or exfiltration. The network operations are strictly targeted at obtaining weather information based on user-provided locations.
  • [PROMPT_INJECTION]: The instructions and scripts do not contain any patterns intended to bypass AI safety guidelines or override agent behavior. The _extract_location function in the Python script safely handles input by trimming and limiting the length of the location string.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 1, 2026, 09:26 AM
Security Audit — agent-trust-hub — weather