Overview

Warden watches over your code by running skills against your changes. Skills are Markdown prompts that define what to look for: security vulnerabilities, correctness bugs, API contract issues, performance problems, or anything else that needs consistent review coverage.

The important split:

Layer	What it controls
warden.toml	Which skills run, which files they see, and how findings are reported.
GitHub workflow	When GitHub starts Warden and which credentials/action inputs are available.
Skill files	The review expertise Warden applies to changed code.

Every analysis run follows the same loop:

1. Identify what changed: files, hunks, directories, or an explicit git range.
2. Match changes against configured triggers.
3. Run the matching skills against the matching code.
4. Report findings with severity, location, and optional fixes.

Skills follow the agentskills.io specification. Warden includes security-review for baseline AppSec coverage and code-review for correctness bugs. Treat them as first passes, then add more skills when your codebase needs deeper coverage.

When to Use Warden

Use Warden when you want consistent review coverage before human review starts, when specialized issues are easy to miss, or when your team needs to enforce patterns specific to your codebase.

Skip Warden for quick typo fixes, formatting-only changes, generated code, or vendored dependencies unless you have a skill for that specific work.

Where It Runs

Warden works in two contexts:

Context	Use it for
Local development	Review changes before pushing.
GitHub Actions	Review pull requests and post findings inline.

Reading Path

QuickstartInstall Warden, authenticate model access, and initialize a repository.ArchitectureFollow changed code through triggers, skill agents, post-processing, and reports.Local ReviewRun branch, staged, range, file, and autofix reviews before pushing.GitHub ReviewsRun Warden automatically on pull requests.Agent SkillHave coding agents run Warden before they commit.SkillsUse built-ins, add remote skills, and write codebase-specific reviews.

Tips

• Use --verbose to see which triggers matched and what Warden is doing.
• Use --fix to apply suggested fixes interactively.
• Use --skill <name> to skip trigger matching and run one skill directly.
• Use --fail-on <level> to override CI failure behavior.
• Use -vv for debug output with token counts and latencies.

Warden also ships as an agent skill so coding agents can run it before they commit.