scholar-review

Perform deep code review on Rust crates by running the scholar-review skill with a crate name argument. It leverages local AI reasoning to analyze architecture, invariants, API ergonomics, performance, and test coverage comprehensively.

You can analyze multiple Rust crates at once using the batch processing feature. This allows efficient, simultaneous AI reasoning across several crates to evaluate API ergonomics, performance, and architectural invariants.

AI code review for Rust does not require an internet connection or send data to external servers. It performs offline analysis using locally available models, ensuring your source code remains entirely private.

Focus AI reasoning on specific aspects like architecture or test coverage by utilizing the selective review feature. This allows you to target particular issues or components within a Rust crate instead of scanning everything.

The primary limitation of using local AI models for Rust code analysis is the requirement to have a capable model pre-installed locally. Deep reasoning quality depends entirely on the local hardware and the selected model's capabilities.

The best way to check algorithmic correctness and invariants in Rust is using AI-driven deep code analysis. This approach evaluates architectural invariants and algorithmic correctness locally without sending proprietary code externally.