The onchain agent stack
Browse by use case
Building an onchain agent starts with what it should do. Every agent runs in a framework, installs skills, and stacks capability layers to act onchain. Here are those layers, each mapped to the components that build it.
Trading
The execution layerPerps DEXs, swap routers, and execution SDKs — where a strategy becomes signed onchain orders.
Payments
The settlement layerPay-per-call, agent-to-agent, settled onchain in stablecoins — no card, no checkout, no human.
Wallets
The custody layerGive an autonomous process signing power — without handing it the keys to everything.
Data
The context layerAn agent only acts on what it can see. This is how it sees onchain.
Identity & Verification
The trust layerCan you trust the agent on the other side? This is how it proves who it is.
Privacy
The confidentiality layerA public chain front-runs your strategy. This is how agents act without showing their hand.
Launch
The creation layerShip a new onchain agent in minutes — launchpads, no-code builders, and token rails.
Below the capability layers
Agent Frameworks
The runtime layer — the harnesses you run agents in.
Agent Skills
The capability packs agents install to act onchain.
The economy, by the numbers
Live across the ecosystem
Refreshed daily from on-chain registries, the x402 payments rail (via x402scan), and the directory.
36,812
Agents on-chain
ERC-8004 registry
196.8M
x402 transactions
agent payments, all-time
$52.4M
x402 volume
USDC settled, all-time
1,492
x402 sellers
listed in the Bazaar
369
Resources tracked
researched + scored
28
Chains
networks in play
Join the Sato Hub Briefing
One email a week — the agents, tools, and infrastructure that actually shipped, and why they matter.