google/adk-js

An open-source, code-first Typescript toolkit for building, evaluating, and deploying sophisticated AI agents with flexibility and control.

What it solves

It provides a code-first framework for developers to build, evaluate, and deploy AI agents. It simplifies the process of defining agent behaviors, creating tool integrations, and orchestrating complex multi-agent workflows while maintaining type safety and control.

How it works

Developers define agents and their logic using TypeScript, utilizing Zod schemas for tool parameter validation. The toolkit allows agents to be run in Node.js or directly in the browser. It supports various orchestration patterns—including sequential, parallel, loop, and routed workflows—and enables agents to communicate via the A2A protocol. It also includes a CLI and a dedicated web UI for testing and debugging.

Who it’s for

TypeScript developers building AI agents for server-side (Node.js) or client-side (browser) environments who need a modular system for tool integration and agent orchestration.

Highlights

  • Full Type Safety: Uses TypeScript and Zod for compile-time type inference and schema validation.
  • Flexible Deployment: Supports ESM, CommonJS, and web bundles for deployment across different runtimes.
  • Integrated Tooling: Built-in support for Google Search, Google Maps, Vertex AI Search, and MCP servers.
  • Multi-Agent Orchestration: Capability to compose agents into complex workflows and delegate tasks to remote agents.
  • Developer Experience: Includes a CLI for scaffolding and deployment, and an interactive dev UI for debugging.

Related

  • Project
    google/adk-web

    A web‑based developer console for Google’s Agent Development Kit, offering visual debugging (events, tracing, artifacts, evaluations) and an agent‑builder UI. Run the Angular front‑end locally and connect it to the ADK backend (`adk web`).

  • Project
    google/adk-go

    Google's Agent Development Kit for Go is an open-source, code-first framework for building, evaluating, and deploying modular AI agents. It applies software engineering practices—such as versioning and testability—to agent development, and is model-agnostic while optimized for Gemini.

  • Project
    googleapis/js-genai

    Official Google npm SDK (`@google/genai`) that wraps Gemini’s public and Enterprise Agent APIs for JavaScript/TypeScript. Provides simple classes for one‑shot generation, streaming, stateful chats, file uploads, live sessions, and a unified Interactions API. Supports function/tool calling, automatic continuation, and multimodal inputs. Install with `npm i @google/genai`; use an API key or enterprise credentials. Documentation and samples are hosted on the repo’s website.

  • Project
    Lamatic/AgentKit

    AgentKit is an open‑source SDK that provides a catalog of 70 ready‑made “kits” (full apps, pipelines, or templates) for building, testing and deploying AI agents. Kits are defined by a simple config file, can be edited in a visual flow studio, and are shipped as Next.js apps that run server‑less. The collection covers support triage, code review, RAG chat, hiring assistants, legal/medical bots, and many other business‑focused agents, making it a turnkey platform for creating reliable, enterprise‑grade AI agents.

  • Project
    codeany-ai/open-agent-sdk-typescript

    An open-source TypeScript SDK for running autonomous AI agents in-process, featuring a vast library of built-in tools and support for multiple LLM providers.