prompt-engineering

Pass

Audited by Gen Agent Trust Hub on Apr 22, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides numerous examples for using the 'infsh' CLI, which is the vendor's own tool (inference-shell), to run tasks and manage AI models. These commands are legitimate and intended for user education.
  • [EXTERNAL_DOWNLOADS]: The documentation references installation instructions and related skills hosted on the vendor's official GitHub repository (inference-sh/skills). These are trusted resources within the context of the skill's authorship.
  • [INDIRECT_PROMPT_INJECTION]: The skill defines several prompt templates that ingest external data (e.g., '[code]', '[article text]'). While this provides an attack surface where malicious input could influence the model's behavior, it is an inherent property of prompt engineering templates and no automated exploitation paths are present.
  • Ingestion points: SKILL.md (Code Review and Content Writing templates)
  • Boundary markers: None explicitly defined in templates
  • Capability inventory: Uses the Bash tool via the 'infsh' CLI
  • Sanitization: None described
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 22, 2026, 04:25 PM
Security Audit — agent-trust-hub — prompt-engineering