AR / FIELD LAB / INTERACTIVE SANDBOX

Try a safe agent action.

Run a local example. Nothing is sent. No real action happens.

  1. DescribeDescribe what the agent wants to do.
  2. PermissionSay whether it has permission.
  3. EvaluateEvaluate the safe path.
  4. Create or blockCreate or block the receipt.
  5. CheckCheck what the receipt proves.
01 / INTENTION

What the agent wants

02 / SAFE PATH

Safe result

READY
Submit a synthetic scenario to evaluate the safe path.
03 / RECEIPT

Receipt identity

AR-PENDING

No receipt generated.

04 / OPTIONAL

Public-safe activity

No activity request is made automatically.

Public-safe aggregate not loaded.
05 / LIMITS

What this proves

Receipt identity, declared mandate linkage, and internal integrity state.

External execution, payment, identity and legal validity remain outside this check.