dowhy
Pass
Audited by Gen Agent Trust Hub on Jun 22, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides instructions and examples for causal inference using the dowhy framework from Microsoft Research. No malicious patterns, prompt injections, or exfiltration attempts were detected.
- [EXTERNAL_DOWNLOADS]: The skill mentions installing well-known, legitimate data science packages including dowhy, networkx, matplotlib, pandas, and scikit-learn from the official Python Package Index (PyPI). All URLs point to official documentation or source code repositories on GitHub.
- [COMMAND_EXECUTION]: Provided code snippets demonstrate standard usage of data analysis libraries for simulating data, building causal graphs, and estimating treatment effects. These operations are local and consistent with the skill's analytical purpose.
Audit Metadata