auth-security-expert

Pass

Audited by Gen Agent Trust Hub on Oct 23, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill's primary objective is to review and refactor code provided by the user, which serves as an ingestion point for untrusted data that could contain malicious instructions.
  • Ingestion points: External code snippets and architecture descriptions provided by the user for security review.
  • Boundary markers: The skill uses XML-like tags (e.g., <instructions>, <identity>) to demarcate its own operational logic.
  • Capability inventory: The skill is configured with powerful tool access, including Bash, Write, Edit, and Read across the filesystem.
  • Sanitization: There are no explicit instructions to sanitize or escape user-provided content before the agent processes it.
  • [COMMAND_EXECUTION]: The instructions include a 'Memory Protocol' section that explicitly directs the agent to execute a shell command (cat .claude/context/memory/learnings.md) at the start of a session. While this is a documented pattern for session persistence in some development environments, it involves direct filesystem interaction based on instruction blocks.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 23, 2026, 06:40 PM