ml-best-practices
Pass
Audited by Gen Agent Trust Hub on Oct 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest and process user-provided datasets and requirements, which creates a vulnerability surface for indirect prompt injection if the input content contains instructions designed to influence agent behavior.
- Ingestion points: Data analysis and machine learning workflows involve reading external schema descriptions, data rows, and task requirements as outlined in the Clustering, Classification, and Regression sections (SKILL.md).
- Boundary markers: The skill does not provide explicit delimiters or instructions for the agent to treat external data as untrusted or to isolate it from the agent's logic.
- Capability inventory: The skill enables the agent to execute code for data transformation, visualization, and model training, which are capabilities that could be abused if triggered by malicious input.
- Sanitization: No sanitization, validation, or filtering steps for external data are specified in the instructions.
Audit Metadata