Zanii turns every move an agent makes into a signed, hash-chained receipt, anchored to a public blockchain and verifiable by anyone. Proof, not logs.
AI agents now touch real systems, calling tools, moving money, accessing customer data. But the only record is a log you could quietly edit. When an auditor, a customer, or a court asks “what did your agent actually do, and who authorized it?”, “trust our database” is not an answer. Zanii makes the answer cryptographic.
Every agent gets a cryptographic identity, delegated by its owner with scoped, expiring authority. An agent can never do more than the human behind it.
Each action becomes an Ed25519-signed, hash-chained receipt in an append-only Merkle log. Deleting, altering, or reordering any receipt is mathematically detectable.
The log's fingerprint is periodically written to Base, so not even Zanii can rewrite history. Verify any proof with the open SDK.
Point the SDK at any function your agent already calls. It signs, hash-chains, and submits the receipt for you. The tool behaves exactly as before.
// every call becomes a provable receipt
import { ZaniiAgent } from '@zanii/sdk'
const zanii = new ZaniiAgent({
serverUrl, agentDid, agentPrivateKey, delegation, apiKey
})
const sendEmail = zanii.wrapTool('email.send', realSendEmail)
// sendEmail works exactly as before —
// and is now cryptographically recorded.Signed receipts attributable to one agent identity. Forgery needs the private key.
Owner-signed, expiring authority with one-click revocation. Least privilege, provable.
Inclusion & consistency proofs (RFC 6962) — the log is provably append-only.
Tree heads anchored to Base, so history is immutable even to the operator.
Self-contained, offline-verifiable exports mapped to EU AI Act & SOC 2 needs.
Two agents co-sign one neutral proof. ERC-8004 compatible identities.
Paste any receipt hash. We fetch it with its Merkle inclusion proof and the signed tree head straight from the public API, the same bundle you could verify offline.
