overseas-trip-planner-skill

Pass

Audited by Gen Agent Trust Hub on Jun 30, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes a local Python script (trip_concierge.py) to process trip parameters and provide structured data for the agent. The test suite (tests/test_trip_concierge.py) also uses subprocess execution to validate the script's behavior. This execution is limited to the skill's own code and does not involve untrusted inputs or shell execution of external binaries.
  • [EXTERNAL_DOWNLOADS]: The skill and its associated scripts do not perform any network operations. The Python script is documented as 'stdlib-only' and analysis confirms it does not import any networking libraries or fetch external data.
  • [PROMPT_INJECTION]: The instructions in SKILL.md are focused on the intended travel concierge persona. There are no patterns suggesting attempts to bypass safety guardrails, extract system prompts, or override agent behavior maliciously.
  • [CREDENTIALS_UNSAFE]: No hardcoded API keys, tokens, or sensitive credentials were found. The skill processes travel preferences and public geographic data (sunset times) without requiring or accessing sensitive user data or system files.
  • [DATA_EXFILTRATION]: All data processing is performed locally. There is no network capability within the script or instructions to send data to external domains.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 30, 2026, 01:49 PM
Security Audit — agent-trust-hub — overseas-trip-planner-skill