unattended-development-v2

Pass

Audited by Gen Agent Trust Hub on Jun 14, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns, obfuscation, or security vulnerabilities were identified in the skill instructions or configuration.
  • [DATA_EXPOSURE]: The skill explicitly instructs the agent to stop and request user input if "real environment credentials" or high-risk decisions (security, costs, etc.) are encountered during the development process, providing a safeguard against unauthorized credential use.
  • [COMMAND_EXECUTION]: The instructions mandate an "anti-simulation" policy, ensuring that the agent does not use mocks or faked logs to bypass real execution and verification steps.
  • [PROMPT_INJECTION]: The skill establishes strict gatekeeping and review stages (using a separate review skill) to ensure that generated outputs are consistent with project requirements and technical designs before proceeding.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 14, 2026, 01:39 PM
Security Audit — agent-trust-hub — unattended-development-v2