Does Warden support remote skills?

Yes, Warden can fetch and sync skills from GitHub repositories, allowing teams to share and version-pin standardized analysis patterns across different projects.

Warden is an event-driven AI agent that analyzes code changes using configurable skills to produce structured reports on bugs, security vulnerabilities, and code quality.

Can I create my own analysis skills?

Yes, you can create custom skills by placing a SKILL.md file in the .warden/skills/ directory containing specific analysis instructions for the agent.

How do I run Warden locally?

You can run it via the CLI using the 'warden' command to analyze uncommitted changes, or target specific files and git references like 'warden main..HEAD'.

Can Warden fix the issues it finds?

Yes, by using the --fix flag in the CLI, Warden can attempt to automatically apply suggested fixes to the findings it identifies during analysis.

Warden Code Analysis

Name: Warden Code Analysis
Author: getsentry

bygetsentry

•

118

•

安全与测试

Analyzes code changes and executes configurable AI skills to generate structured security and bug reports.

Warden is an event-driven AI agent designed to streamline code reviews by automatically identifying bugs, security vulnerabilities, and style issues. By integrating directly with the Claude Code environment and local CLI, it allows developers to run sophisticated analysis skills—either custom-built or fetched from remote repositories—against uncommitted changes or specific git refs. Whether you need to enforce strict security policies on critical files or automate routine linting tasks, Warden provides a flexible framework via warden.toml to ensure high code quality and security before every commit.

主要功能

01Local CLI for analyzing uncommitted changes and specific git ranges

02Parallel execution and flexible model selection for optimized performance

03118 GitHub stars

04Support for custom and remote skills fetched directly from GitHub

05Configurable warden.toml for automated triggers and file filtering

06Structured reporting with severity levels and automated fix suggestions

使用场景

01Performing automated security audits on sensitive authentication and payment modules

02Running pre-commit bug hunts to catch logic errors before pushing code

03Standardizing code review processes across teams using shared remote skills

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add getsentry/dotagents warden-skill

For use in Claude.ai and ChatGPT

Download Skill