novel-evaluator

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied story texts or outlines, creating a surface where malicious instructions embedded in the stories could influence agent behavior.
  • Ingestion points: The skill accepts story texts or outlines as input within the SKILL.md instructions.
  • Boundary markers: The skill lacks explicit delimiters or instructions to ignore embedded commands in the story text, though it specifies constraints on evaluation.
  • Capability inventory: The skill is authorized to use the Read tool.
  • Sanitization: There is no mention of sanitization or filtering of the input story content.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 06:49 PM
Security Audit — agent-trust-hub — novel-evaluator