Build a Code Graph from Your Repository with Vibgrate Graph

Vibgrate Graph turns your source tree into a structural graph artifact — nodes for functions, classes, and modules, edges for calls and imports. That graph is the foundation for every other query: vg show, vg ask, vg impact, vg path, vg tree, and the map-level insights. This tutorial builds your first graph.

Prerequisites

• A source repository you can scan
• Node.js installed

Steps

1. Install the Vibgrate CLI

npm i -g @vibgrate/cli

Prefer no install? Use npx @vibgrate/cli for any command.

2. Initialize Vibgrate in the project

vg init

This creates the .vibgrate directory and config file so artifacts have a home.

3. Build the code graph

vg build

vg build maps your source code into a graph artifact incrementally. The first run indexes the whole tree; later runs only reprocess what changed. The artifact powers all downstream queries — fully offline after this point.

4. Check graph status

vg status

vg status reports graph freshness, node and edge counts, staleness, and the resolver rungs used. It compares the committed graph against your current working tree so you can see whether the graph is up to date.

5. Rebuild incrementally after edits

After you change code, run:

vg build

The incremental build keeps the graph current without a full re-index, so it stays fast on large repositories.

Verification

Run vg status and confirm node and edge counts are non-zero and that staleness reports the graph matches your working tree. A healthy build means downstream commands like vg show and vg ask will return results immediately.

Next Steps

• Explain a single node with vg show
• Ask the graph a question with vg ask
• Map architectural hubs with vg hubs