paper-writer
Pass
Audited by Gen Agent Trust Hub on Aug 15, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill interacts with reputable scientific platforms such as PubMed, CrossRef, OpenAlex, and Semantic Scholar to conduct literature searches and verify the authenticity of citations.
- [COMMAND_EXECUTION]: The skill utilizes Bash and Python scripts to automate word counting, perform statistical tests (e.g., t-tests, survival analysis), and compile manuscript sections. These operations use standard research libraries and tools like pandoc.
- [SAFE]: The skill incorporates significant security and integrity measures, including advice on protecting patient data and a dedicated red-teaming process designed to identify and correct potential AI-generated hallucinations or faked results before submission.
Audit Metadata