defense-in-depth

Fail

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: HIGHDYNAMIC_CONTEXT_INJECTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [DYNAMIC_CONTEXT_INJECTION]: The skill utilizes the !command syntax designed to execute shell commands at load time to populate the agent's context.
  • Evidence in SKILL.md: !cat CHANGELOG.md | head -n 20 executes a file read command.
  • Evidence in SKILL.md: !ls -R executes a recursive directory listing command.
  • [INDIRECT_PROMPT_INJECTION]: The skill's primary purpose is to process and validate local code and workspace files. This creates a vulnerability surface where malicious instructions embedded in the analyzed data (e.g., comments in code or documentation) could influence the agent's behavior.
  • The skill lacks explicit instructions for sanitizing or escaping the content it reads from the workspace.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Sep 9, 2026, 03:52 AM
Security Audit — agent-trust-hub — defense-in-depth