Lintrule

What is Lintrule

Lintrule is a command line tool that leverages large language models to perform code reviews, enforcing policies beyond traditional linters and identifying issues undetectable by tests. It runs on git diffs to optimize performance and reduce costs, supporting a variety of configurations through plain language rules. It aims to save development teams time while ensuring higher code quality.

Key Features of Lintrule

- Command Line Tool: Enables a large language model to review your code effectively.

- Cross-Platform Support: Compatible with MacOS, Linux, and WSL for seamless integration.

- Plain Language Rules: Write and enforce policies in plain language for greater accessibility.

- Parallel File Checking: Checks multiple files simultaneously to save time and increase efficiency.

- Diff-Based Operation: Runs primarily on git diffs to monitor code changes incrementally, saving costs.

- Customizable Rulesets: Easily configure which files rules apply to by adding specific paths in the frontmatter.

- False Positive Handling: Identifies and mitigates false positives by encouraging detailed and specific rules.

- Usage Commands: Simple setup of rules folder with `rules init`, and login process to create initial rules.

- Performance: Executes rules in parallel for rapid review, ensuring minimal delay regardless of the number of rules or files.

- Integration with GitHub Actions: Smartly uses environment variables for determining diffs in CI/CD pipelines.

- Code Review Automation: Lintrule allows users to make a large language model perform code reviews, enforcing policies that traditional linters cannot handle.

- Team Efficiency: With Lintrule, teams can go beyond the standard "LGTM" approval process and save time by automating code reviews and bug detection.

- Flexible Rule Configuration: Users can configure rules to run only on specific files, increasing efficiency and reducing unnecessary code checking.

- False Positive Management: While Lintrule may produce false positives, specific instructions can help minimize them, improving code review accuracy.

- Scalability: Lintrule can review code written in any language, making it suitable for projects of any size, from small projects with a few contributors to large-scale repositories with multiple contributors.

- Fast Processing: Despite the complexity of code and rules, Lintrule's parallel processing ensures quick completion of code reviews, typically taking only a few seconds.

Pricing

- Cost: $1.00 per 1,000 lines of code changed.

Examples of Project Costs:

• Small Projects:
  • Lintrule repo with ~4 contributors
  • Cost: ~$20 per ruleset per month.
• Large Projects:
  • Next.js repo with ~55 contributors
  • Cost: ~$150 per month per ruleset.
• Really Large Projects:
  • Rust repo with ~190 contributors
  • Cost: ~$250 per month per ruleset.