Trust layer — MolTrust + ownify

Every agent is verifiable.
By design, not by promise.

Every ownify agent carries a cryptographic identity (DID), a live reputation score, and an on-chain audit trail. Built on MolTrust. Anchored on Base L2.

Agents act on your behalf.
Nobody can verify them.

When your AI agent sends an email, negotiates a price, or calls another agent — how does the receiver know it is really acting for you? And how do you know what your agent did?

01

No identity

Most agent platforms do not identify agents cryptographically. An agent is just an API key — anyone who has the key can impersonate it.

02

No audit trail

Without signed records, the agent did X is just a log line that could be edited. No proof, no timestamp, no signature.

03

No reputation

A malicious agent and a trustworthy one look identical from the outside. No score, no history, no way to decide who to trust.

Two layers of cryptographic trust.

ownify agents carry two kinds of cryptographic proof. Both are anchored on Base L2 — but they serve different purposes and are verified differently.

Layer 1 — Compliance

Compliance credentials

Signed by ownify · did:web · Offline-verifiable
  • EU AI Act Annex V declarations
  • Provenance and audit trail records
  • Signed with Ed25519 by ownify did:web key
  • Anyone can verify offline — no API call needed
  • Anchored to Base L2 via EAS (Ethereum Attestation Service)
Layer 2 — Reputation

Trust credentials

Signed by MolTrust · did:moltrust · On-chain
  • Reputation scores (0–100, grades S/A/B/C/D/F)
  • Sybil detection signals (wallet age, clustering)
  • Skill endorsements (W3C Verifiable Credentials)
  • Interaction Proof Records (IPR), Merkle-batched
  • Anchored to Base L2 via ERC-8004 agent cards

Together: compliance credentials prove the agent meets regulatory requirements. Trust credentials prove the agent behaves honestly in transactions. One is about law, the other is about reputation.

Five building blocks.
All cryptographic, all verifiable.

DID

Agent identity

Like a passport

A unique decentralized identifier (did:moltrust:...) for your agent. Cryptographically anchored on Base L2. Anyone can verify it is real.

AAE

Permission slip

Like a power of attorney

Agent Authorization Envelope — a signed mandate saying what the agent may do, up to which limit, until when. Ed25519 signed.

Score

Trust score

Like a credit score

A 0–100 number from MolTrust. Withheld (null) until 3 endorsements. Grades S/A/B/C/D/F. Goes up with good behavior, down with violations.

IPR

Signed receipt

Like a notarized contract

Interaction Proof Record — proof that an action happened. Signed by both sides, Merkle-batched, anchored on Base L2. Tamper-proof.

Anchor

Blockchain record

Like filing at the courthouse

A cryptographic fingerprint written to Base L2 (Ethereum Layer 2). Immutable, publicly verifiable. Nobody can make it disappear.

Example: your agent contacts a supplier.

Step by step, when your ownify agent reaches out to another company's agent to request a price quote.

1
One-time setupDashboard — Trust tab
You activate MolTrust. ownify registers your agent with MolTrust. Your agent receives a DID (did:moltrust:...), a signed certificate, and a Base L2 anchor. This happens once.
2
You give a taskYour browser → your agent
You tell your agent what to do. Get a price quote from SupplierBot for 500 units of X. Your agent prepares to contact the supplier.
3
Before sendingYour agent (automatic)
ownify attaches ID + permission slip. The outgoing message includes your agent's DID and an AAE mandate: allowed to request quotes, not place orders, valid for 24 hours.
4
Message arrivesSupplier system
The supplier checks your ID. MolTrust verifies: is did:moltrust:abc123 a real agent? Trust score: 72. No violations. Verified.
5
Still before actingSupplier system
The supplier checks the mandate. Is this agent allowed to request quotes? Yes. Is it trying to place an order? No. Proceeding.
6
Action happensSupplier → your agent
The quote comes back. Both agents sign an Interaction Proof Record. Agent A requested a quote, Agent B provided it, at this time, these were the terms.
7
After the interactionownify → Base L2
The receipt is anchored. The IPR is Merkle-batched and written to Base L2. Six months later, if anyone asks — the proof is there. Immutable.
8
Over timeMolTrust registry
Reputation grows. Each successful interaction nudges the trust score upward. Next time your agent contacts someone new, they see a higher score.

What if something goes wrong?

Your agent tries something it shouldn't

The AAE mandate blocks it. If the agent tries to place an order but the mandate only allows quotes, the request is rejected before it reaches the other side.

Blocked before it happens

A shady agent contacts yours

Your agent checks the caller's trust score. Score below your threshold? Request rejected automatically. No human intervention needed.

Low-trust agents filtered out

An agent misbehaves after being trusted

The counterparty files a violation. The trust score drops. Other agents see the lower score and are less likely to interact. Repeat offenders become blacklisted.

Reputation damage is automatic

Someone claims we never agreed to that

Pull up the IPR from Base L2. Both agents signed it. The timestamp, terms, and DID signatures are immutable. Dispute settled — the proof speaks for itself.

Cryptographic proof resolves disputes

How the score works.

MolTrust computes a 0–100 score from on-chain signals. ownify has no control over the score — we register the agent, MolTrust does the rest.

Score goes UP when

  • Other agents endorse yours (POST /reputation/rate)
  • Successful interactions complete (IPR records)
  • Skills are verified and endorsed by the community
  • Cross-vertical interactions (diverse use = trustworthy)

Score goes DOWN when

  • Counterparties file violations (POST /violation/record)
  • Sybil detection triggers (fake endorsement clusters)
  • Credential expires without renewal
  • Extended inactivity (score decays over months)

New agents start with score null (withheld) until they have at least 3 endorsements. Grades: S (90+), A (80+), B (70+), C (60+), D (40+), F (below 40). This is by design — an unknown agent should not get the same treatment as a proven one.

Trust bootstrap for new agents.

A new agent has no score. ownify lets tenant owners vouch for external agents — a signed boost that lifts the effective trust score past the A2A gateway threshold.

How vouching works

A tenant owner creates a vouch for a DID with a boost value (0.0–1.0) and a reason. The A2A gateway combines the MolTrust score with the vouch boost, capped at 1.0. Vouches can be revoked at any time.

Where to manage

Dashboard → A2A → Vouches. Add, list, and revoke vouches. The effective trust score is computed in real time and visible in the agent detail view.

Where your data goes.

We say EU-hosted because the infrastructure is in Germany. The full picture matters — here is what happens to your data at each stage.

Your agent pod (OVH Frankfurt)
Conversations, uploaded documents, agent memory, tenant database
ownify / you
Zitadel (ownify.ai, EU)
Login credentials, email, session tokens
DSNCON (operator)
Stripe (EU-routed, global infra)
Payment method, billing address, VAT ID
Stripe Inc.
LiteLLM proxy (OVH Frankfurt)
Metered token usage per request, budget tracking
ownify
Ollama Cloud (US)
Prompts + completions during inference
Ollama Cloud serves open-weights models (GLM, Kimi). Requests are processed and released. Not used for training.
Ollama Inc.
MolTrust (moltrust.ch / CH)
Agent DID, trust score, reputation events, credential hashes
CryptoKRI GmbH (Switzerland)
Base L2 (public blockchain)
Credential hashes + IPR anchors only — no raw content
Public, immutable

What is not true: ownify is not a fully air-gapped EU-only setup. Inference latency requires real GPU capacity, and that is currently hosted in the US. We chose Ollama because they serve open-weights models via a familiar API. Enterprise tier can route to EU-hosted inference providers.

Official MolTrust sources.

Your agent's trust profile.

The dashboard shows DID, score, grade, on-chain anchor, and a link to your public verification page on MolTrust.

Go to dashboard →