How it works

The loop has six steps. Here is each one, with the command that does it.

Step 1
Push
Write a package manifest with your instructions, skills, hooks and configs, then `wl push --package <path>` uploads it as a draft package version. The registry runs its content scan first.
Step 2
Promote and review
Once the content scan passes, `wl publish` moves the draft to pending_review. A maintainer approves or rejects it in the dashboard, after seeing its blast radius.
Step 3
Assign
A workspace declares which packages render into which targets for which providers. `wl init` or `wl bind <manifest>` writes the binding, .loom/binding.toml, in your repository.
Step 4
Render
`wl pull --dry-run` shows the render plan; `wl pull` writes each provider’s files and records them in the state file. Trust-sensitive assets wait for approval.
Step 5
Check drift
`wl status` and `wl diff` show each file’s ownership state. `wl sync` exits 1 on drift, so it can gate CI, and `wl restore` puts Workloom-owned files back.
Step 6
Govern
Roles decide who can publish and approve, workspace revisions make every change reversible with `wl rollback`, and the audit log records who changed what.

Stop copying AI config between repositories

Publish your instructions, skills and hooks once, assign them to workspaces, and let CI tell you when anything drifts.