AgentKit

AgentKit is a framework for creating and orchestrating AI Agents, from single model inference calls to multi-agent systems that use tools. Designed with orchestration at it’s core, AgentKit enables developers to build, test, and deploy reliable AI applications at scale.

• Overview
• Installation
• Documentation
• Examples

Overview

A networked agent:

// Create a network of agents with separate tasks and instructions
// to solve a specific task.
const network = createNetwork({
  agents: [navigator, classifier, summarizer],
  defaultModel: openai({ model: "gpt-4o", step }),
});

const input = `Classify then summarize the latest 10 blog posts
  on https://www.deeplearning.ai/blog/`;

const result = await network.run(input, ({ network }) => {
  // Use an agent which figures out the specific agent to call
  // based off of the network's history.
  return defaultRoutingAgent;
});

A simple agent:

const writer = createAgent({
  name: "writer",
  system:
    "You are an expert writer.  You write readable, concise, simple content.",
  model: openai({ model: "gpt-4o", step }),
});

const { output } = await writer.run(
  "Describe the ideas behind the given input into clear topics, and explain any insight: " +
    `<content>${content}</content>`,
);

Installation

Agent kit requires the Inngest TypeScript SDK as a dependency. You can install both via npm or similar:

npm install @inngest/agent-kit inngest

Documentation

The full Agent kit documentation is available here. You can also jump to specific guides and references:

• Agents and Tools
• Network, state, and routing

Examples

See Agent kit in action in fully functioning example projects:

• Test Writing Network - A ready-to-deploy Next.js demo using the Workflow Kit, Supabase, and OpenAI to power some AI content workflows.

License

Apache 2.0