See it work — in your world
AI agents are getting authority across the business. Pick an area below and try what an agent might do — every result is real, computed live by the open-source Scopebond gateway, with a signed receipt as proof.
Pick what the agent tries to do:
Signed by the gateway (Ed25519). Anyone can verify it against the published key at /v1/attester — no trust in this server required.
Out-of-policy actions are blocked in real time — fail closed. The bad action never happens.
Every action, allowed or blocked, is a tamper-evident receipt. Audit- and court-ready.
Back a third-party agent with a deposit and recover on a provable violation.
Open-source and free to self-host — or move to hosted/enterprise when you want it off your plate.
Each area is a policy evaluated live by the same gateway. Example:
curl -sX POST https://try.scopebond.com/v1/case/finance/evaluate -H 'content-type: application/json' -d '{"intent":{"action_type":"payout.create","asset":"USDC","amount":500000}}'
Inspect: /v1/status · /v1/attester · /.well-known/jwks.json
Run your own: npx scopebond-gateway ./policy.json · GitHub