feature-engineering-kit
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [Indirect Prompt Injection] (LOW): The skill ingests untrusted data from CSV files which could be manipulated to include malicious instructions for downstream agent tasks.
- Ingestion points:
scripts/feature_engineering.pyreads user-provided files viapd.read_csv. - Boundary markers: Absent; the tool does not provide delimiters or warnings to ignore embedded data instructions.
- Capability inventory: The skill has file-writing capabilities via
to_csvinscripts/feature_engineering.py. - Sanitization: Absent; data is processed and transformed without escaping or validating the content against instruction-like strings.
- [Data Exposure & Exfiltration] (SAFE): No hardcoded credentials, sensitive file path access, or unauthorized network operations were identified. The script only reads and writes files specified via command-line arguments.
- [Remote Code Execution] (SAFE): The skill uses standard data science libraries (pandas, scikit-learn) and does not utilize
eval(),exec(), or any subprocess calls to execute remote or dynamically generated code.
Audit Metadata