autonomous-agents

Pass

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: LOWNO_CODE
Full Analysis
  • [NO_CODE] (SAFE): The skill consists entirely of markdown documentation and metadata. There are no executable scripts, tool definitions, or configuration files that could trigger side effects.
  • [Prompt Injection] (SAFE): No malicious instructions or bypass attempts were found. The content focuses on agent reliability and safety patterns (e.g., 'Least privilege principle', 'Agent guardrails').
  • [External Downloads] (SAFE): No remote resources, package managers, or external scripts are referenced for download or execution.
  • [Indirect Prompt Injection] (SAFE): The skill does not define any tools that process external or untrusted data; it is an informational resource for the agent's reasoning context.
Audit Metadata
Risk Level
LOW
Analyzed
Feb 16, 2026, 07:23 AM