informed-captain

Pass

Audited by Gen Agent Trust Hub on Apr 17, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill contains only markdown-based instructional content and JSON-based test cases. No executable scripts, binaries, or active components were found.
  • [PROMPT_INJECTION]: No patterns associated with prompt injection, safety bypass, or role-play manipulation were detected in the instructions or test prompts.
  • [DATA_EXFILTRATION]: The skill does not contain any network-facing tools, hardcoded credentials, or instructions to access sensitive local files.
  • [COMMAND_EXECUTION]: No shell commands, privilege escalation attempts, or persistence mechanisms are present within the skill files.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 17, 2026, 04:54 AM
Security Audit — agent-trust-hub — informed-captain