ml4t-causal-identification
Pass
Audited by Gen Agent Trust Hub on Sep 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is entirely instructional, providing best practices for causal identification using Directed Acyclic Graphs (DAGs) and refutation tests. It recommends using industry-standard libraries such as DoWhy, EconML, and scikit-learn for these tasks.
- [SAFE]: No evidence of prompt injection, data exfiltration, or obfuscation was found. The instructions are focused on statistical validity and preventing common errors like conditioning on colliders or mediators.
- [SAFE]: The code snippets provided are benign examples of linear regression and causal estimation workflows. There are no attempts to execute arbitrary shell commands, access sensitive files, or establish unauthorized network connections.
Audit Metadata