Ephemeral Environments

To validate agent-built changes against a running system, Cat-Factory can provision ephemeral preview environments on demand and tear them down when the run finishes.

What they're for

Tests that need a live instance, such as integration tests, Playwright end-to-end runs, and acceptance checks, get an isolated environment spun up just for that run, then cleaned up automatically.

How it works

You register a preview environment provider via a declarative HTTP manifest that points at your provisioning and cleanup endpoints. During a run:

The deployer agent calls your provider endpoints to spin up an isolated environment.
The tester (and playwright) agents run against the preview instance.
The environment is cleaned up automatically on run completion or timeout.

Run starts
   └─ deployer agent → provision environment (your HTTP API)
        └─ tester / playwright → run against preview
             └─ run completes or times out → environment torn down

Registering a provider

You describe your provider declaratively with a manifest. There are no per-provider presets and no per-org code:

Field	Purpose
Base URL + auth scheme	Where to reach your management API and how to authenticate to it.
provision template	Called to spin up an environment.
status template	Polled until the environment is ready.
teardown template	Called to tear the environment down.
Response mapping	Maps your API's responses onto a canonical environment handle (e.g. its live URL).

Credentials are referenced by logical key, never embedded. You supply the values at registration, where they're stored encrypted at rest. The manifest's structure is documented in Integration Manifests, and it's enabled through the Environment provider manifest feature toggle. See Configuration → Feature Toggles.

Automatic cleanup

Environments are removed on completion or timeout, so a stuck run won't leave preview infrastructure (and cost) running indefinitely.

That covers deployment and operations. For internals, see the Reference.