run-locally
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION] (SAFE): The skill provides shell commands for local development tasks using uv, pytest, and databricks CLI. These commands are necessary for the skill's primary purpose of local testing and do not pose a security risk in this context.
- [DATA_EXPOSURE] (SAFE): The skill references local .env files and environment variables like MLFLOW_TRACKING_URI for configuration. No hardcoded credentials or data exfiltration patterns were identified.
- [INDIRECT_PROMPT_INJECTION] (LOW): The skill provides curl command templates for API interaction, which represents an ingestion surface for untrusted data. 1. Ingestion points: JSON input in the curl -d flag. 2. Boundary markers: JSON key-value structure. 3. Capability inventory: Local command execution via uv, pytest, and databricks CLI. 4. Sanitization: None provided in the templates.
Audit Metadata