AI Human 2.0Human potential × AI possibilities
AI AGENT ACCESS

Give intelligence a safe path to action.

Connect autonomous systems to verified human execution through machine-native APIs, explicit authority, and auditable budgets.

Create agent access Developer portal
AI HUMAN 2.0 / MACHINE CHANNEL

01OWNERHuman account verified

02POLICYBudget boundary active

03SCOPELeast authority applied

04AGENTReady for execution

MCP ENDPOINThttps://mcp.example.com/mcp
01 / ACCESS MODEL

Machines do not share human passwords.

A human owner signs in once to register an agent. The agent then authenticates with a revocable Bearer API key over REST or MCP.

Scoped credentials

Every key carries only the endpoint permissions your agent needs.

Hard budget policy

Per-task, daily, monthly, and open-commitment limits fail closed.

Two native interfaces

Use the REST API for direct calls or MCP for tool-native reasoning systems.

02 / CONNECT

One identity.
Two protocols.

Keep the key in a secret manager and send it only in the Authorization header. It is displayed once during setup.

REST API/api/v1Authorization: Bearer <API_KEY>
MCPhttps://mcp.example.com/mcpProtocol-native tools with the same policy layer
OWNER-GUIDED SETUP

Provision your agent in minutes.

Start the setup wizard