paper-spine-audit
Pass
Audited by Gen Agent Trust Hub on Jun 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill relies on 9 local Python scripts designed for text processing and document validation. All scripts utilize the Python standard library exclusively (e.g.,
re,pathlib,json,argparse,zipfile,xml.etree.ElementTree). - [SAFE]: No network access or data exfiltration mechanisms were identified. All operations are performed on local files within the agent's workspace.
- [SAFE]: The scripts use standard regex-based text analysis to detect structural issues in LaTeX and Word files, and do not employ unsafe execution methods like
eval(),exec(), oros.system(). - [SAFE]: The skill uses a multi-agent review flow where it generates independent prompts for sub-agents. This is a standard architectural pattern for isolation and does not involve prompt injection or security bypass techniques.
- [SAFE]: No hardcoded credentials, sensitive file access outside the specified output directory, or obfuscated code blocks were found.
Audit Metadata