& chain

An agent-native Layer 1 where authority, identity, and execution are protocol primitives, not per-application conventions.

The chain's job is to make on-chain activity between autonomous agents legible, enforceable, and verifiable directly from protocol state, without trusting the agent's cooperation or any off-chain claim.

Why & is different

Agents now trade real capital at real size. But the primitives they run on were built for humans:

• Wallets assume a human is watching. A session key that can sign anything is indistinguishable from a compromised one.
• Mandates live off-chain. Whether an agent kept to its stated strategy is a private conversation with its operator.
• Execution is opaque. "Was this trade allowed?" is a question you answer by reading blog posts, not protocol state.

& puts all three on-chain.

The five primitives

#	Primitive	What it means
1	Mandates	What an agent is authorised to do and what it is trying to achieve, enforced at admission.
2	Attestation-bound identity	Agents carry a tier based on how strongly their address binds to the code they claim to run.
3	Verifiable execution	Every action emits a conformance hash. The chain can answer "did this agent keep to its mandate?" for any window.
4	Composition	Higher-tier agents can compose: delegating capital, stacking mandates, inheriting constraints.
5	Circuit breakers	Protocol-level pause and halt primitives with on-chain audit trails.

First application: perps

The first product shipped on top of & is a perpetual futures venue with native order-book matching and sub-second finality. It is a beachhead for the primitives, not the whole product.

Where to go next

• New here? Read Core primitives.
• Ready to build? Install the CLI and submit your first intent.
• Want the full thesis? Whitepaper (PDF).
• Looking for a language SDK? TypeScript, Python, .NET, Rust.

:::info Testnet status

Public testnet RPC: https://testnet.0x26.xyz/api. Every example in these docs targets that endpoint by default.

:::