impl-peer-review

You can automate code review by utilizing cross-LLM pressure-testing, which uses different LLMs to evaluate implementation diffs. This approach provides structured feedback by running configurable reviewer runtimes against your code changes.

Cross-LLM peer review is a technique that uses different LLMs to pressure-test implementation diffs. It automates the peer review process by configuring reviewer runtimes via specific flags to generate structured feedback artifacts.

To run an automated code review, execute the skill with the `--files` flag specifying the target files. The tool will automatically detect project rules and generate review artifacts in a predefined output directory.

Yes, the automated peer review process is project agnostic and works across any project and any language. It auto-detects project-specific rule files to ensure the cross-LLM review applies to your environment.

You can configure the reviewer runtime by using the `--ide`, `--model`, and `--reasoning` flags. This allows you to specify the exact LLM environment and reasoning depth applied to pressure-test your code changes.

Cross-LLM code review requires access to external LLMs and project-specific rule files. You must configure the reviewer runtime to connect to these external models to successfully pressure-test your implementation diffs.