Search code, repositories, users, issues, pull requests...
A declarative language for defining AI agents, their capabilities, and tools. Think of ADL as "OpenAPI for AI Agents" - a standardized specification that enables consistent agent definition, documentation, and code generation across platforms. ADL simplifies agent development by generating production-ready code from a single manifest. Build faster. Deploy smarter. Stay consistent.
ADL (Agent Definition Language) is a unified, declarative specification language designed to address the growing complexity and fragmentation in the AI agent ecosystem. Just as OpenAPI revolutionized API development by providing a standard way to describe REST services, ADL does the same for AI agents.
Inspired by the pain points observed in Agent-to-Agent (A2A) communication and Model Context Protocol (MCP) implementations, ADL was created to solve the fundamental challenges of agent interoperability and standardization.
Following Domain Driven Design principles, ADL allows you to model your agent's domain logic, capabilities, and behaviors in a vendor-agnostic way. As the AI industry experiences explosive growth with new tools, frameworks, and vendors emerging daily, ADL provides a standardized approach to define, configure, and deploy AI agents across different platforms and environments while maintaining clear domain boundaries and consistent terminology.
