skills/mo7yw4ng/openarc/setup-source/Gen Agent Trust Hub

setup-source

Pass

Audited by Gen Agent Trust Hub on Jul 24, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill guides the agent to assist users in building a local integration while maintaining a strong security posture.
  • [CREDENTIALS_UNSAFE]: The instructions explicitly mitigate credential exposure by commanding the agent to 'Keep credentials out of the code' and instead 'pull them from the environment.'
  • [COMMAND_EXECUTION]: While the skill results in the generation of executable Python code, it includes clear boundaries to 'Keep it narrow' and avoid dangerous capabilities like bulk-scraping or multi-account evasion.
  • [DATA_EXFILTRATION]: The skill handles user-provided network captures which may contain sensitive data, but instructs the agent to use this information only for structural reproduction of API calls on the user's local system.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 24, 2026, 06:18 AM
Security Audit — agent-trust-hub — setup-source