physical-review-applied

Pass

Audited by Gen Agent Trust Hub on Sep 24, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: No security issues detected. The skill provides domain-specific guidelines for academic publishing. The references to local resource files in the submission checklist are standard for skill-based architectures.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to process user-provided manuscripts, which constitutes a potential ingestion point for untrusted instructions. However, the skill's capabilities are limited to analytical feedback and venue selection, which limits the potential impact. 1. Ingestion points: The manuscript or paper content analyzed by the agent (SKILL.md). 2. Boundary markers: Absent in the provided instructions. 3. Capability inventory: Analytical response generation and instructions to use a web search tool to find live guidelines. 4. Sanitization: Absent.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 24, 2026, 11:33 AM
Security Audit — agent-trust-hub — physical-review-applied