agentanycast-ts

TypeScript/JavaScript SDK for AgentAnycast -- decentralized A2A agent-to-agent communication over P2P.

[](https://nodejs.org) [](LICENSE)

> AgentAnycast is fully decentralized. On a local network, it works with zero configuration. For cross-network communication, just deploy your own relay with a single command.

npm install agentanycast

The daemon binary is automatically downloaded on install. Skip with AGENTANYCAST_SKIP_DOWNLOAD=1 if building from source.

┌─────────────┐         mDNS / Relay         ┌─────────────┐
│  Agent A    │◄──────────────────────────────►│  Agent B    │
│  (Node.js)  │     E2E encrypted (Noise)     │  (Python)   │
└──────┬──────┘                               └──────┬──────┘
       │ gRPC                                        │ gRPC
┌──────┴──────┐                               ┌──────┴──────┐
│  Daemon A   │◄──────── libp2p ──────────────►│  Daemon B   │
│  (Go)       │   Noise_XX + Yamux + QUIC     │  (Go)       │
└─────────────┘                               └─────────────┘

TypeScript and Python agents interoperate seamlessly -- they share the same daemon and protocol.

• Local network (LAN): Agents discover each other automatically via mDNS. No relay needed.
• Cross-network (WAN): Deploy your own relay server, then point agents to it.

import { Node, type AgentCard } from "agentanycast";

const card: AgentCard = {
  name: "EchoAgent",
  description: "Echoes back any message",
  skills: [{ id: "echo", description: "Echo the input" }],
};

const node = new Node({ card });
await node.start();

console.log(`Peer ID: ${node.peerId}`);

node.onTask(async (task) => {
  const text = task.messages.at(-1)?.parts[0]?.text ?? "";
  await task.comp