create-command

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_CONTEXT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill acts as a command generator that builds new command files from user-supplied descriptions and requirements, creating an ingestion surface for untrusted data.\n
  • Ingestion points: User requirements and descriptions provided during the interactive interview process defined in SKILL.md.\n
  • Boundary markers: The skill lacks explicit delimiters or instructions to ignore commands potentially embedded within the user's requirements.\n
  • Capability inventory: The skill performs directory listings and file reading (using the @ prefix) to study existing patterns and generates new command files on the local filesystem.\n
  • Sanitization: No sanitization, validation, or escaping of user-provided content is performed before it is interpolated into the generated command templates.\n- [DYNAMIC_CONTEXT_INJECTION]: The skill provides documentation and examples for executing shell commands at skill load time using the ! prefix. These instances are located within educational markdown blocks and guides for command creation, and do not represent active or unauthorized command triggers within the skill's own functional context.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 01:10 AM
Security Audit — agent-trust-hub — create-command