extracting-characters-and-backgrounds

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes an external file, 'story.md', as its primary input without implementing boundary markers or sanitization. This creates a surface where instructions embedded within the story could influence the agent's behavior.
  • Ingestion points: The skill reads content from 'story.md'.
  • Boundary markers: There are no instructions to the agent to distinguish between the story narrative and potential instructions within the file.
  • Capability inventory: The skill performs file read/write operations (characters.json and backgrounds.json).
  • Sanitization: No filtering or validation of the input text is specified.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 02:15 AM
Security Audit — agent-trust-hub — extracting-characters-and-backgrounds