Dependency Management

Xavier maintains best-practice skills for your key dependencies. Instead of re-fetching documentation every session, your AI agents get pre-distilled guidance on how to use each package correctly.

What's a dependency-skill?

A self-contained knowledge bundle stored in ~/.xavier/skills/<dep-name>/ containing:

• Best practices distilled from official documentation
• Links to full/official docs
• API patterns relevant to your codebase

When you or your agents work with that package, Xavier has the context ready — no doc-fetching required.

Commands

Add a dependency

/xavier add-dep <package-name>

Xavier fetches the package's documentation, distills the most important best practices, and writes a skill file to the vault.

/xavier add-dep zod

This creates ~/.xavier/skills/zod/ with a skill file containing Zod's validation patterns, type inference best practices, and links to the official docs.

Remove a dependency

/xavier remove-dep <package-name>

Removes the skill directory from the vault.

/xavier remove-dep moment

Update stale dependencies

/xavier deps-update

Reads your current lockfile versions, compares them against existing dependency-skills, and regenerates any that are out of date.

Xavier can't passively detect when you update a dependency — deps-update is the explicit trigger. Run it after significant dependency bumps.

How skills are used

Dependency-skills are part of your vault's knowledge graph. When Xavier performs operations like code review or task planning, it can reference these skills to provide better guidance.

For example, if you have a Zod skill and you're reviewing code that adds input validation, the reviewer knows Zod's recommended patterns and can flag anti-patterns specific to that library.

Ecosystem support

Currently supports Node.js projects (package.json / lockfiles). Other ecosystems (Rust, Python, Go) are planned for future versions.