Everything you need
to control AI agents.

AUVEN gives agents you create or connect an identity, daily limits, owner approval for supported requests, and post-quantum signed activity.

AUVENCONTROLLED REQUEST / 01Illustrative flow
ACONNECTED AGENTResearch assistantACTIVE
⌁POLICY GATEWAYDaily limit checked✓
◇SIGNED ACTIVITYCheckpoint createdML-DSA-65
IDENTITYCONTROLPROOF
$AUVEN ON SOLANA

$AUVEN token

Token mintDXf5UDjzYFEizcroAHx1jhtFLf6EEuWcSokgNfAhpump

From identity to
verifiable activity.

AUVEN sits between connected agents and supported tools. Explore the three parts that make their activity accountable.

Know who is acting.

Create a text agent in your workspace or register one you already run. Each gets its own gateway credential and an ML-DSA-65 signed passport.

RResearch assistantConnected agentActive
Agent IDagt_demo_01
Gateway credentialIssued ·•••••••
AGENT PASSPORTML-DSA-65
Identity signedVerified ✓

From agent identity
to accountable action.

Create a text agent here or connect one from your application. The gateway applies your rules to supported requests and records the outcome.

01 / CREATE OR CONNECT

Give it an identity.

Define the agent's purpose and save its own gateway credential.

◈
02 / CONTROL

Decide what can pass.

Set a daily limit and keep the agent active or stopped. The gateway rejects requests that fail a check.

⌁
03 / VERIFY

Inspect the evidence.

Review decisions, export signed checkpoints, and verify the signature independently.

▧

Post-quantum proof,
built into the record.

ML-DSA-65 signs agent passports and activity checkpoints. A verifier can check that an exported record is authentic and unchanged without trusting the dashboard alone.

Digital signatures are verified offchain. Blockchain proof commitments are a separate operation and require a configured chain signer.

Connect what you already build.

Register an external agent, save its credential, then connect AUVEN to Codex or Claude Code with MCP. Calls made through the AUVEN tool pass your agent's policy checks and appear in Activity.

MCP setup guide
REQUEST FLOW
YOUR AGENT→AUVEN GATEWAY→OPENAI
identity ✓    policy ✓    activity ✓

Give every agent
a clear place to start.

Create an agent in your workspace or connect one you already use.

Register an agent identity

Name an agent you already run. MCP or API setup comes next.

The agent credential appears once after registration, with setup commands you can copy.

WORKSPACE ACCESS