journal-of-climate

Pass

Audited by Gen Agent Trust Hub on Sep 24, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides structured instructions and heuristics for evaluating academic paper fit for the Journal of Climate. It contains no executable scripts or system commands.
  • [SAFE]: No evidence of prompt injection, obfuscation, or malicious intent was found in the instructions or metadata.
  • [SAFE]: The skill does not perform network operations or access sensitive local files beyond its own resource directory.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-submitted manuscript content to provide feedback. While this represents a data ingestion surface, the skill lacks capabilities such as shell execution or network exfiltration that could be leveraged by malicious input.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 24, 2026, 11:29 AM
Security Audit — agent-trust-hub — journal-of-climate