The Agent Skills Directory

[SAFE]: The skill provides architectural guidelines, design patterns, and crate recommendations for machine learning development in Rust. No malicious code, data exfiltration patterns, or obfuscation were detected.
[INDIRECT_PROMPT_INJECTION]: The skill describes patterns for ingesting external numeric data (f32 tensors) for model inference.
Ingestion points: The predict and batch_predict functions in SKILL.md accept external data inputs.
Boundary markers: Not present in the provided code snippets.
Capability inventory: No dangerous capabilities (e.g., file writing, network operations, or subprocess execution) are utilized in the examples.
Sanitization: Not present. However, because the input type is limited to numeric tensors and the skill lacks exploitable capabilities, the configuration is safe.

domain-ml