series-bible

Pass

Audited by Gen Agent Trust Hub on Aug 10, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill instructions are focused on creative writing continuity. No patterns for system prompt extraction, safety bypass, or role-play injection were detected.
  • [DATA_EXFILTRATION]: The skill uses local file storage for its operations. There are no network requests, hardcoded credentials, or attempts to access sensitive system files.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided story data. As it lacks network or high-privilege command capabilities, the risk of cross-context instruction injection is minimal and consistent with standard agent functionality.
  • [COMMAND_EXECUTION]: The skill does not define or invoke any shell commands or external scripts.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 10, 2026, 07:41 PM
Security Audit — agent-trust-hub — series-bible