ma-peer-review
Pass
Audited by Gen Agent Trust Hub on Jun 19, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill's primary logic is implemented through transparent Python scripts that perform statistical parsing and data transformation. These scripts utilize the Python standard library (csv, json, re) to process research artifacts and generate reviews.
- [SAFE]: No evidence of data exfiltration or credential harvesting was found. The skill does not perform network operations or access sensitive system directories.
- [SAFE]: External URL references in the documentation point to established academic and research resources, such as the University of Bern's CINeMA tool for network meta-analysis, which is a well-known service in the medical research community.
- [SAFE]: Command execution is limited to running the included local scripts via the 'uv run' command, which is standard behavior for the execution environment and does not involve arbitrary or untrusted command injection.
Audit Metadata