the-bmj

Pass

Audited by Gen Agent Trust Hub on Sep 24, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill provides structured guidance for editorial evaluation, including topic fit, methodological bars, and house style. It does not contain any executable code, scripts, or network exfiltration logic.
  • [INDIRECT_PROMPT_INJECTION]: The skill's primary function is to analyze clinical medicine manuscripts, which are untrusted external data sources.
  • Ingestion points: Clinical manuscripts and research metadata provided to the agent for evaluation.
  • Boundary markers: None explicitly defined in the skill instructions.
  • Capability inventory: The skill references reading local resource files and performing web searches for official journal guidelines.
  • Sanitization: No specific sanitization or filtering of manuscript content is mentioned. While this creates a theoretical injection surface, there are no instructions to treat the manuscript content as executable instructions or to bypass safety guardrails based on the input document.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 24, 2026, 11:33 AM
Security Audit — agent-trust-hub — the-bmj