template-skill

Pass

Audited by Gen Agent Trust Hub on Nov 5, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill functions as a template and does not contain any executable code, network requests, or access to sensitive files or credentials.
  • [PROMPT_INJECTION]: The skill uses repetitive instructional emphasis (e.g., 'IMPORTANT MUST ATTENTION') to enforce specific reasoning behaviors and task-tracking protocols. These are standard prompt engineering techniques intended to improve agent reliability and do not attempt to bypass safety filters or ethical guidelines.
  • [INDIRECT_PROMPT_INJECTION]: The instructions reference external protocol files (e.g., '.claude/skills/shared/protocols/evidence-based-reasoning.md'). These are modular local references for reasoning standards and do not involve untrusted data ingestion.
  • [METADATA_POISONING]: The metadata fields contain placeholder text (e.g., '[Category] Use when defining...') which is appropriate for a template and contains no deceptive or malicious instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 5, 2026, 06:42 PM