Description
Perch is an open-source semantic linter designed to enhance code quality by reviewing pull requests for bugs, security issues, and custom rules. Powered by SystemOne decision models, it provides a comprehensive analysis of code changes, ensuring that potential issues are identified and addressed before integration. Perch supports a wide range of programming languages, including Python, TypeScript, JavaScript, Go, Rust, Java, Kotlin, C, C++, C#, Swift, Ruby, and PHP, among others.
Read More
The tool operates by parsing the repository with tree-sitter, mapping each method's callers and callees, and running predefined rules alongside its built-in bug and security scans. Users can write custom rules in perch.yaml, such as ensuring S3 buckets are never public, which are applied to every pull request. Perch comments directly on the pull request, highlighting any flagged code.
Perch's pricing model is straightforward, charging approximately 20¢ per pull request, with 10¢ per 1,000 rule checks. Cached checks are free, and the first $5 of usage is complimentary, with no credit card required to start. This makes it an affordable option for developers and teams looking to maintain high code quality without incurring significant costs.
The tool is particularly useful for code owners and developers who want to enforce specific coding standards and security practices across their projects. By integrating with GitHub and running in GitLab CI, Perch offers flexibility in deployment, catering to various development environments.
Perch does not train on user code, ensuring privacy and security. Findings are deleted 30 days after each scan, aligning with its privacy policy. This focus on security and privacy makes Perch a reliable choice for developers concerned about data protection.
- CategoryAI Code Review & Testing
- Pricing ModelFreemium
- DeploymentCloud
- PlatformsGitHub, GitLab
Perch's Core Features
Open-source semantic linter
SystemOne decision models
Custom rule support
20¢ per pull request
10¢ per 1,000 rule checks
Supports multiple programming languages
GitHub and GitLab CI integration
Privacy-focused with no code training
How to use Perch?
Install: Set up Node 22 and git
Configure: Write custom rules in perch.yaml
Use: Run Perch on pull requests
Optimize: Review flagged issues and refine rules
Perch's Use Cases
- Code Quality Assurance
- Custom Rule Enforcement
- Security Compliance
- Cost-Effective Code Review
- Multi-Language Support
