How-to · CLI

How to get productive with the vem CLI on day one

January 21, 20269 min read

Learn the core CLI flows to create tasks, capture decisions, and commit verified memory alongside your code.

Initialize and align your repo

Start by scaffolding the .vem directory and ensuring your team agrees on how memory artifacts are stored and reviewed.

Use the CLI to generate task objects so that every item of work has a durable ID, status, and timestamp.

Keep the task index short and current so it can act as the first stop for agents and reviewers.

Initialize memory artifacts

vem init

Add your first task

vem task add --title "Set up memory layer" --priority high

Record significant architectural or workflow choices as ADRs. This keeps the why next to the what.

Because decisions are committed with code, reviewers can validate them alongside the implementation.

Short, consistent ADRs beat long essays. Focus on the decision, its context, and the tradeoffs you accepted.

The CLI helps keep changelog entries, task updates, and context packs in sync with the code you ship.

The result is a predictable, searchable history that agents can query without guesswork.

Treat every memory update like a product change: review it, merge it, and let the indexer verify it.

Review what changed

vem status

Commit memory with your code

vem commit -m "Add billing org switcher"

Send a pending snapshot

vem push

Add at least one task, one decision, and one changelog entry during the first sprint.

Make sure your team can retrieve those artifacts through search before inviting agents to use them.