KURAL
Commands

Overview

CLI reference for every kural command

Each command is one page: synopsis, flags, an example, and when to reach for it. Conceptual background (why the command exists, what it computes) stays in the matching Concepts page — the reference pages link to it rather than repeat it.

At a glance

CommandWhen to use
snapshotBootstrap the repo — parse source, embed every unit, write the local database.
briefBefore writing new code — show what already exists in the area.
placeBefore writing new code — rank directories by fit to find its home.
auditAfter writing — list specific structural findings on the updated snapshot.
advisePrioritize findings into an ordered action plan for one directory.
scoreTrack aggregate health over time — fit, uniqueness, children fit/uniqueness.

Typical flow

Run kural snapshot generate src on the repo — and again whenever source changes.

Before adding code: use kural brief to see what exists, and kural place to find its home.

After editing: run kural audit for findings, then kural advise for a prioritized action plan.

Use kural score to track aggregate health over time — pin snapshots and compare.

All analysis commands read from the snapshot database — they do not re-embed. Regenerate the snapshot after meaningful code changes so findings stay current.

On this page