Lucid Agents
The protocol-agnostic multi-runtime framework for building and monetizing AI agents.
Build, deploy, and monetize autonomous AI agents with typed entrypoints, on-chain identity, and built-in payment infrastructure.
Quick start
Get your first agent running in minutes:
bunx @lucid-agents/cli my-agent
cd my-agent
bun run devIntroduction
Learn what Lucid Agents is and why you should use it
Installation
Install packages and set up your environment
Quickstart
Build your first agent in 5 minutes
Core concepts
Agents
Understand agent configuration and the extension system
Entrypoints
Define typed API capabilities with Zod schemas
Build with agents
AI Router
OpenAI-compatible x402 router with permit sessions and streaming responses
Skills
Install reusable skill files with a curl command and use them in your AI client
Autonomous agents
Use agents to create and evolve full applications on top of Lucid
Key features
- Type-safe APIs - Define inputs/outputs with Zod schemas, get automatic validation and TypeScript inference
- Framework flexibility - Deploy on Hono, TanStack Start, Express, or Next.js
- Accept payments - x402 protocol for USDC payments on Base or Solana
- On-chain identity - ERC-8004 standard for verifiable agent reputation
- Agent-to-Agent - A2A protocol for agent discovery and communication
- Real-time streaming - Server-Sent Events for LLM responses and long operations