crafting-instructions

Pass

Audited by Gen Agent Trust Hub on Apr 25, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill's primary function is to serve as a meta-instructional guide for crafting prompts, projects, and skills. All instructions follow industry best practices for instruction engineering and do not attempt to subvert agent safety or behavioral guidelines.
  • [COMMAND_EXECUTION]: The reference documentation (creating-skills.md) contains standard shell commands such as mkdir, zip, and git which are provided as educational material to help users structure and package their own skill directories. These operations are benign and aligned with the skill's stated purpose.
  • [DATA_EXFILTRATION]: Analysis found no evidence of hardcoded credentials, sensitive file access, or unauthorized network requests. References to platform-specific storage paths like /mnt/user-data/outputs/ and computer:/// URIs are standard for managing file outputs in the intended developer environment.
  • [PROMPT_INJECTION]: The skill contains instructional examples of what to avoid in prompt design, including how to handle complex tasks without attempting to bypass safety filters or core system constraints.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 25, 2026, 12:19 PM