hawker-discover-skill

Pass

Audited by Gen Agent Trust Hub on Jun 25, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill invokes a Python script from a sibling skill (cdc-voucher-locator-skill) using subprocess.run. The invocation uses a hardcoded relative path and passes arguments as a structured list, avoiding shell-based command injection risks.
  • [EXTERNAL_DOWNLOADS]: The skill retrieves real-time data and datasets from official Singapore government APIs (api.data.gov.sg and api-open.data.gov.sg) and the OneMap geocoding service. These connections are made to well-known technology and public service domains for the legitimate purpose of data retrieval.
  • [CREDENTIALS_UNSAFE]: The skill documentation and .env.example file correctly guide the user to manage API keys using environment variables and gitignored files, which is an industry-standard security practice.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 25, 2026, 04:52 AM
Security Audit — agent-trust-hub — hawker-discover-skill