Agentic Economy, Coordination and Orchestration Protocols — Overview and Discussion
AI agents are rapidly evolving from experimental tools into autonomous economic actors.
They already write code, operate infrastructure, trade assets, run services, and interact with APIs. The next step is an agentic economy — where autonomous agents interact with each other across networks, coordinate work, exchange value, and discover services without human mediation.
But the protocol layer for the agentic internet is still forming.
If millions — and eventually billions — of AI agents are going to operate across the internet and blockchains, we will need shared standards and protocols that allow them to interoperate reliably and securely.
Some important pieces already exist across different ecosystems.
Ethereum ecosystem
• ERC-8001 — Agent Coordination Framework
Intent-based coordination primitive allowing agents to propose, accept, and execute tasks collaboratively.
• ERC-8004 — Trustless Agents
Registry framework for agent identity, reputation, and validation.
• EIP-712 — Structured Data Signing
Machine-readable signed intents and permissions.
• ERC-4361 — Sign-In With Ethereum (SIWE)
Authentication standard for agents interacting with web services.
• ENS — Ethereum Name Service
Human-readable identity and service discovery layer.
• EAS — Ethereum Attestation Service
Portable credentials, reputation, and verifiable claims.
Solana ecosystem
• Solana Actions & Blinks
Standardized way to distribute executable blockchain actions via URLs.
• Solana Pay
Machine-readable payment protocol suitable for automated agents.
• Solana Name Service (SNS)
Naming and identity layer.
Agent-native protocols emerging outside blockchains
• A2A (Agent-to-Agent Protocol)
Standardized communication between autonomous agents.
• ANS — Agent Name Service
DNS-like identity and discovery system for AI agents.
• x402 protocol
HTTP-native micropayment layer enabling agents to pay for APIs and services.
• ACP — Agent Commerce Protocol
Marketplace protocol enabling autonomous agent commerce.
• MCP — Model Context Protocol
A protocol to connect AI models to external tools, data sources, and services
OpenClaw and the rise of agent orchestration
In parallel, new tools like OpenClaw are making it easier to deploy and orchestrate autonomous AI agents in real-world environments.
OpenClaw enables agents to:
• interact with the filesystem and operating system
• execute complex workflows
• manage memory and context
• call APIs and services
• coordinate multiple sub-agents
• automate research, development, and operations tasks
This represents an important shift — from single LLM tools toward multi-agent systems operating as autonomous operators.
But this also raises new infrastructure questions.
If thousands of OpenClaw-style agents begin interacting with each other, we will need standards for:
• agent identity and registries
• capability discovery
• secure delegation of tasks
• agent-to-agent messaging
• reputation systems
• payment rails between agents
• sandboxing and safety mechanisms
• orchestration of multi-agent workflows
• memory and context sharing between agents
Today, many of these primitives exist only partially.
Key questions we will explore
• What protocols already exist that could power the agentic economy?
• What critical primitives are still missing?
• How should OpenClaw-like agents discover and trust each other?
• What orchestration standards should exist for multi-agent systems?
• Should coordination happen on-chain, off-chain, or hybrid?
• What payment rails should autonomous agents use?
• How should reputation systems work for autonomous software?
• What new EIPs, SIMDs, or open protocols should be created?
Format
This is an overview of existing protocols and an open discussion on what is missing.
It is an open discussion club where participants collaboratively explore the protocol stack needed for the agentic internet.
• informal discussion circle
• presentation on what protocols already exist in the industry
• participants share ideas, research, and experiments
• collaborative exploration of open problems
Who should join
• AI agent builders
• protocol designers
• Web3 developers
• infrastructure engineers
• researchers
• founders exploring autonomous systems
• anyone curious about the future of the agentic internet
Outcome
A shared understanding of the protocol architecture required for the agentic economy, and potentially the seeds of new standards.
Come with ideas. Leave with better questions.