Vibecode at full speed
without getting wrecked.
Rollcage snapshots your production database before an AI agent can break it, then asks you first.
- Claude Code
- Cursor
- Codex
- Copilot
- Devin
- Antigravity
- Gemini CLI
- Cline
- opencode
- Amp
- Droid
- Qwen Code
Works with the agent you already use.
Twelve of them, one set of rules. Postgres, MySQL and SQLite.
Where each hook livesFrom risky command to safe, in one beat.
Clean up the old tables before launch.
The users table looks like a leftover from the old auth. I'll drop it.
psql "$DATABASE_URL" -c "DROP TABLE users"- Command
DROP TABLE users- Aimed at
- production (Supabase)
- Found in
DATABASE_URLin.env- Rule
sql-drop-table
users48,213 rowsorders120,904 rowsinvoices9,377 rows
Encrypted on this computer, then sent to Cloudflare R2.rc_20261121_0942
psql "$DATABASE_URL" -c "DROP TABLE users"This would DROP TABLE users on production (Supabase). Snapshot saved: rc_20261121_0942. Allow it?
$ rollcage undo
The snapshot from just before DROP TABLE users, 2 minutes ago.
Restoring users (48,213 rows) into a new database...
Restored into acme_restore_0942. Nothing was overwritten.
$ rollcage copy-back --preview
Would put back 48,213 rows. Rows added since stay.
Everything an agent can wreck, covered.
Before you launch: test your live app.
Launch check looks at your running site the way a stranger would, for the mistakes that get vibe-coded apps wrecked, and says how to fix each one. It runs on your computer, only reads, and only checks sites you've shown are yours.
- Public .env files
- Keys in your JavaScript
- Open Supabase tables
- Public .git folders
- Admin pages with no login
- Debug pages
- HTTPS and headers
61 requests, all reads, in 8.4 s.
- CriticalYour .env file is public
/.env: DATABASE_URL=…, STRIPE_SECRET_KEY=…Fix: take it out of public/, then change every key that was in it.
- CriticalSupabase table profiles is readable by anyone
profiles: 1,204 rows, 9 columnsFix: turn on row-level security for profiles, with a policy for each user's own row.
- LowNo Content-Security-Policy
/Fix: add a Content-Security-Policy header listing where scripts may load from.
About 4 ms.
You'll never notice it.
Safe commands go straight through, without touching the network. Rollcage only slows down when it's saving your data.
- A typical check
- 3.9 ms
- Agents it works with
- 12
- Rows of your data it sees
- 0
Every close call, in one place.
See what your agents tried, and approve from your phone the close calls an agent can't stop to ask you about. The dashboard gets metadata only: never connection strings, keys or a row of your data.
Sign in to the dashboardDROP TABLE legacy_ordersproductionApproveDenyTRUNCATE sessionsproductionYou allowed itrm -rf uploadsfilesYou said no
Questions.
A small command-line tool that sits between your AI coding agent and your production database. It checks every command the agent runs, takes an encrypted snapshot before anything destructive reaches production, and asks you before it runs.
Rollcage is in private testing right now. Join the waitlist and you'll get one email when it opens. Installing will take one line, and a settings page walks you through the rest.
Claude Code, Cursor, Codex, GitHub Copilot (the CLI, and its agent in VS Code), Devin (formerly Windsurf), Antigravity (the app, its IDE and the agy CLI that replaced Gemini CLI), Gemini CLI (still around with paid API keys), Cline, opencode, Amp, Droid (Factory) and Qwen Code. Databases: Postgres (Supabase, Neon, Railway, RDS and any other Postgres you can reach with a connection string), MySQL or MariaDB, and SQLite.
No. Safe commands are checked in about 4 milliseconds without any network calls. The only pause is when a destructive command is aimed at production and Rollcage is saving a snapshot.
Wherever you choose: Cloudflare R2, AWS S3, DigitalOcean Spaces, any S3-compatible storage, or a folder on your computer. They're encrypted before they leave your machine, with a key that only you have.
Say yes. Rollcage isn't there to stop you working. It's there to make sure there's a way back. The snapshot is already saved, and every decision lands in the activity log.
No. Rollcage runs on your computer. Snapshots are encrypted on your machine and go straight to the storage you picked. If you connect the optional dashboard, it gets metadata only: database names and hosts, what Rollcage caught, snapshot ids and sizes, and table names with their row counts. Never connection strings, keys, or any row of your data.
Yes. Type any command into the box at the top of this page or into the Command X-ray, or build your rollcage.toml in the config builder and try commands against it. Both run Rollcage's real engine, compiled to WebAssembly, in your browser: the same code the hook runs, so you get the answer Rollcage would give in a project set up like that. In the playground you can change the project itself: add a rule of your own, protect a folder, change what a script runs. Nothing you type is sent anywhere.
Yes, with Launch check. Start it from the dashboard, or run rollcage scan with your site's address, and it looks for the mistakes that get vibe-coded apps wrecked: a public .env file or .git folder, secret keys in your JavaScript, Supabase tables anyone can read, admin pages that open without a login, missing HTTPS and security headers. You get a score out of 100, and each finding with how to fix it. It runs on your computer and only reads: no forms, no logins, at most 150 requests. It checks sites on your own computer or network, or ones you've shown are yours with a small file or a DNS record. Keys it finds are never kept or sent anywhere: it shows only their kind and first four characters. How Launch check works.
Run rollcage undo. It finds the snapshot Rollcage took just before that command, shows which tables were dropped or emptied since, and brings them back into a new database. Then rollcage copy-back --preview shows exactly what copying back would do to today's data, and rollcage copy-back puts the missing rows back into production once you confirm, keeping rows added since.
Not quietly. Changing Rollcage's settings or hooks, replacing the Rollcage program, reading its key, deleting snapshots, or starting another agent with its hooks off always asks you first, however it's attempted. Your own allow rules can't switch that off.
Run rollcage explain '<the command>'. It shows what Rollcage would do with it and why: which rule matched, which database it's aimed at, and whether a snapshot comes first. Nothing is run. The Command X-ray is made from exactly that.
Go fast. Keep your data.
Rollcage is in private testing. Join the waitlist and we'll email you when it's ready for your project.
One email when Rollcage opens. We never share your address.