container-security

Pass

Audited by Gen Agent Trust Hub on Jun 17, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill consists of instructional content and checklists for security auditing. It does not include any executable scripts or dangerous commands.
  • [PROMPT_INJECTION]: The skill contains a safety notice that explicitly warns the agent to treat external file content as untrusted data, mitigating indirect prompt injection risks.
  • [DATA_EXPOSURE]: The guidelines help identify secrets in target files but do not contain or exfiltrate any sensitive information themselves.
  • [COMMAND_EXECUTION]: The skill is configured with restricted tools (Read, Grep, Glob), preventing unauthorized system modifications or network access.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 17, 2026, 04:53 PM
Security Audit — agent-trust-hub — container-security