depmap

Pass

Audited by Gen Agent Trust Hub on May 28, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill downloads cancer research datasets (CSV files) from well-known scientific resources including the DepMap portal (depmap.org) and Figshare. These operations are consistent with the skill's stated purpose of oncology target validation.
  • [REMOTE_CODE_EXECUTION]: No remote code execution or dynamic evaluation of untrusted input was detected. The download functions are restricted to writing data chunks to local files.
  • [DATA_EXFILTRATION]: Network activity is limited to GET requests targeting official scientific APIs and data repositories. There is no evidence of unauthorized access to local sensitive files or credential harvesting.
  • [PROMPT_INJECTION]: The instructions focus entirely on data analysis workflows and do not contain any patterns intended to bypass AI safety guidelines or override system behavior.
  • [COMMAND_EXECUTION]: The skill uses standard data processing libraries (Pandas, NumPy, SciPy) and does not invoke shell commands or OS-level sub-processes.
Audit Metadata
Risk Level
SAFE
Analyzed
May 28, 2026, 03:41 PM
Security Audit — agent-trust-hub — depmap