Unlocking the Power of the Dependency Graph

The main focus of Rui Adorno's talk is on unlocking the power of the dependency graph and demonstrating how to leverage it using the Vault client.

Rui Adorno is a founder engineer at Vault and has previously worked at NPM Inc., GitHub, and Google. He has extensive experience in the Node.js package management space.

Dependency Selector Syntax (DSS) is a query language based on CSS selectors that allows users to query and manage the dependency graph of a JavaScript project.

DSS helps in managing dependencies by allowing users to query the dependency graph using CSS-like selectors, enabling them to filter and select packages based on various attributes like name, version, author, and security metadata.

VLT introduces two phases for package installation: Phase one involves downloading and extracting packages without executing lifecycle scripts, and phase two involves building selected scripts using dependency selector syntax for targeting.

VLT improves security by blocking all install scripts by default, allowing users to inspect the dependency graph and check for security issues before any code runs, thus preventing potentially malicious scripts from executing automatically.

Yes, DSS can be expanded to query across multiple projects using the host local selector, allowing users to target all configured projects in their system.

AI agents can benefit from a queryable dependency graph by receiving structured data instead of crawling files, resulting in faster and more efficient access to machine-readable output and rich metadata.

Users can use VLT's pkg command to add new properties to package.json files, such as adding dependencies or scripts, across all or selected workspaces within a project.

A dependency graph in JavaScript is a data structure that represents the dependencies of a project. It is created when a package manager resolves dependencies during installation.