The right level of automation for an AI agent is 99%. The right level of access is read and write everything. The remaining 1% is review and publish. That last 1% stays human.
Cursor already does this for code. You open a folder, hit Cmd+K or turn on agent mode, and see the diff before anything goes anywhere. The thing Cursor can't reach is your CMS. Scratch is the connector that puts your content into a folder Cursor can open.
You connect Scratch to wherever your content lives. Shopify, WordPress, Notion, HubSpot, the others. It downloads everything as a folder of files on your laptop. You open that folder in Cursor and tell it what you want done. Rewrite the meta descriptions on these 400 posts so each one is under 158 characters and ends with a call to action. Cursor does it.
Then Scratch shows you every change as a per-record diff against the original. You do the 1%: approve what ships. Scratch publishes only those records back to your CMS, one at a time. The records you didn't approve never leave your laptop.
The question you're actually asking is whether you can trust Cursor with your live site. You can. Publishing is the thing Cursor doesn't get to do.
Where you drive Cursor from
Cursor gives you two natural ways to edit a Scratch folder. Both work on the same files on disk.
Cmd+K is for one record at a time. Open a Scratch file, hit Cmd+K, tell Cursor what you want. Best when you're working out a prompt or nudging a single record that needs special attention.
Agent mode is for the whole folder. Tell Cursor what to do across every record. Tone passes, length caps, FAQ generation, tag normalization. Cursor handles the file walk; Scratch surfaces every change for review.
You bring your own Cursor. Scratch holds no Cursor credentials and runs no model. You sign into Cursor the way you already do, on whichever model you've pointed it at.
The things you're probably worried about
Will Cursor publish anything to my live site? No. Cursor only reads and writes files in your Scratch folder. Scratch is the only thing that talks to your CMS, and it only publishes what you've clicked approve on, one record at a time.
Can I undo if something ships I didn't want? Yes. Every published record is reversible from the Scratch app, per row. The original sits next to the rewritten one until you tell Scratch which one stays.
What about my prices, SKUs, member data, anything sensitive? You lock those fields when you set up the connector, and Scratch can't write them back even if Cursor tries. On top of that, you can run Python validators that reject any diff that touches a field you've flagged. Bad rewrites get filtered out before you have to look at them.
Does anyone see my content differently because it went through Cursor? No. Cursor sends prompts to whichever model you've configured (Claude, GPT, your local model). Scratch doesn't change anything about that data path. Same as if you opened any other folder in Cursor.
How is this different from a Cursor extension that hits my CMS? An extension that talks to your CMS API gives Cursor the publish button. Scratch doesn't. Cursor gets the same access an extension would give it, but the publish step stays in your hands. That's the whole difference.
Can I do this on hundreds of records at once? Yes, that's the use case. Agent mode is built for catalog-scale work.
Do I need to be a developer? If you're already using Cursor, you are. If you're not, the Claude desktop app is the more direct route. Same Scratch loop, more familiar interface.
Browse the skills below for prompts that work end-to-end with Cursor.