camera-ready

Pass

Audited by Gen Agent Trust Hub on Aug 24, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill instructions and configuration exhibit no malicious patterns or unauthorized access attempts. All allowed tools are consistent with the task of processing LaTeX documents and managing academic citations.
  • [INDIRECT_PROMPT_INJECTION]: The skill identifies external reviews as a data source for document edits, creating a potential surface for indirect instructions.
  • Ingestion points: External reviewer comments saved to correspondence folders in Phase 0.
  • Boundary markers: No specific delimiters are defined to separate reviewer input from system instructions.
  • Capability inventory: The skill utilizes Write, Edit, and Bash tools for document compilation and metadata management.
  • Sanitization: The skill requires manual verification of all external citations and author sign-off for math-related changes.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 24, 2026, 06:17 AM
Security Audit — agent-trust-hub — camera-ready