prompt-engineering

Pass

Audited by Gen Agent Trust Hub on Mar 12, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill content is purely educational, offering techniques and templates for improving AI model outputs.
  • [COMMAND_EXECUTION]: The skill includes numerous examples using the infsh CLI tool (e.g., infsh app run). These are standard usage patterns for the vendor's platform and do not involve arbitrary or malicious command execution.
  • [EXTERNAL_DOWNLOADS]: The skill references several external resources and other skills within the inference-sh ecosystem using npx skills add. These references point to trusted vendor repositories and well-known AI services (OpenRouter, Google, Fal.ai) for legitimate model access.
  • [PROMPT_INJECTION]: While the skill provides templates that interpolate user data (e.g., [code], [article text]), which is a common surface for indirect prompt injection, this is inherent to the educational purpose of a prompt engineering guide. No malicious or bypass-oriented instructions were found within the skill itself.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 12, 2026, 06:36 PM