PLAYGROUND
Try it before your users do.
A sandbox in your dashboard where you simulate conversations, test agent invokes, and preview how your manifest behaves — no deploy required.
Why this mode
One Shelli. Three callers.
Safe sandbox
Exercise real shell behavior against your configured actions in an isolated session — playground chat does not collide with your embedded shell on the same browser.
See what callers will see
Preview data views, consent cards, and handoffs exactly as humans and agents will experience them.
Iterate on the manifest
Adjust permissions and actions, push a new version, test again immediately. Roll back in one command if something drifts.
Flow
How it works
Define actions once in duct.config.ts — this is how that contract plays out for this caller.
Define your manifest
Sign up, connect your product, and map your actions in duct.config.ts.
Open the playground
Launch the sandbox from your dashboard for any shell.
Run test turns
Simulate human chat, agent invokes, and cross-shell delegation — inspect responses in real time.
Ship with confidence
When it works in the playground, embed the widget and open the invoke API.