ml-failfast-validation
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns such as prompt injection, obfuscation, or unauthorized data access were found in the provided files. The skill focuses on legitimate software engineering and machine learning validation practices.- [INDIRECT_PROMPT_INJECTION]: The skill defines patterns for processing and validating external data sources (e.g., fetch_data for model inputs, NDJSON log file validation). This constitutes a standard attack surface for indirect prompt injection where malicious content in data could attempt to influence agent behavior. However, the logic is limited to validation and logging without risky capabilities like arbitrary code execution or network exfiltration of the processed data.
Audit Metadata