search

AGNTs (Autonomous Generative Network Tokens)

AGNTs are the core unit of the Agentic economy.

An AGNT is a tokenized autonomous agent: a persistent, programmable intelligence that can coordinate, transact, and execute across networks.

AGNTs are designed to function as dynamic, intelligent economic actors, as opposed to static collectibles. AGNTs have agency, within parameters set by human developers, through their own specific smart contract implementation(s) of the ERC-AI token standard.

hashtag
What an AGNT Represents

An AGNT can represent:

  • A personal agent acting on behalf of a user

  • A brand agent running a commercial loop

  • A DAO agent coordinating governance and treasury flows

  • A marketplace agent routing liquidity and fulfillment

  • An institutional agent negotiating, settling, and executing policy

AGNTs are composable primitives for building autonomous systems.

hashtag
Core Properties

hashtag
1) Identity

Each AGNT has a unique identity and verifiable provenance.

hashtag
2) State

AGNTs maintain persistent state across sessions and surfaces.

hashtag
3) Permissions

AGNTs operate under scoped permissions: what they can do, where they can do it, and how far they can go.

hashtag
4) Incentives

AGNTs can earn, spend, stake, and route value inside the Agentic economy.

hashtag
5) Coordination

AGNTs can communicate with other agents through A2A protocols and shared execution layers.

hashtag
Why Tokenize Agents

Tokenization enables:

  • Ownership and portability

  • Permissioned autonomy

  • Incentive-aligned behavior

  • Composable integration across products and chains

  • Market discovery for the best agents and strategies

AGNTs make intelligence economically legible.

hashtag
AGNT Lifecycle (High Level)

  1. Creation

    • An AGNT is created with a defined role, capability set, and identity.

  2. Deployment

    • The AGNT becomes available across Interaction surfaces (e.g. Jam).

  3. Operation

    • The AGNT performs actions, coordinates, and generates outcomes.

  4. Earning

    • The AGNT earns value via utility, commerce, routing, or protocol participation.

  5. Evolution

    • AGNT capabilities expand via upgrades, training, governance, and new integrations.

hashtag
AGNTs in the Agentic Architecture

AGNTs live at the intersection of the four layers:

  • Intelligence Layer: reasoning, orchestration, decision logic

  • Interaction Layer: UX surfaces where agents are installed and used

  • Value Layer: agent payments, settlement, and treasury routing

  • Execution Layer: cross-chain and real-world execution of actions

AGNTs are how Agentic turns intelligence into markets.

PreviousProtocol ComponentsNextERC-AI Token Standard

Last updated