Command Development

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • Prompt Injection (LOW): The skill documentation describes a framework for interpolating untrusted user data into agent prompts through arguments ($1, $ARGUMENTS) and file references (@path).
  • Ingestion points: Arguments ($1, $2) and file content references (@) are interpolated directly into command prompts across all examples in examples/simple-commands.md.
  • Boundary markers: Examples largely lack explicit delimiters (like XML tags or clear separators) to isolate user-provided arguments from the command's instructions.
  • Capability inventory: The documented system supports high-privilege capabilities including file modification (Write, Edit) and arbitrary command execution via Bash (e.g., npm, kubectl, git).
  • Sanitization: While the README.md mentions "Validation patterns" as an enhancement, the provided references and examples do not include specific sanitization logic for user-provided strings before interpolation.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 08:36 PM