optimization-ml-hybrid
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill documentation and code snippets provide legitimate architectural patterns for hybrid AI-optimization systems using standard Python libraries.
- [SAFE]: No malicious behavior, such as credential theft, remote code execution, or obfuscation, was found in the provided file.
- [INDIRECT_PROMPT_INJECTION]: The skill defines functions that ingest external data, such as demand features and problem instances, into ML models and optimization solvers. Ingestion points: SKILL.md (functions like train_ml and optimize_with_forecast). Boundary markers: None. Capability inventory: ML prediction and optimization solving via PuLP, with no network, shell, or filesystem access. Sanitization: None.
Audit Metadata