functions

Zavu Functions

Zavu Functions = serverless TypeScript on Zavu Cloud + a declarative framework for AI agents.

import { defineAgent, defineTool } from "@zavudev/functions"

defineAgent({
  senderId: process.env.SENDER_ID!,
  name: "Bella",
  provider: "zavu",
  model: "openai/gpt-4o-mini",
  prompt: "You are Bella, host at the restaurant. Be brief.",
})

defineTool({
  name: "check_availability",
  description: "Get free reservation slots for a date.",
  parameters: {
    type: "object",
    properties: { date: { type: "string" }, partySize: { type: "number" } },
    required: ["date", "partySize"],
  },
  handler: async ({ date, partySize }) => {
    return { available: true, slots: ["19:00", "21:00"] }
  },
})

That's a full agent + tool. npx zavudev deploy reconciles the live state.

Project shape

Decide this before you write anything, because restructuring later is the expensive kind of rework.

my-agent/
├── index.ts        entrypoint: defineAgent + defineTool live here
├── lib/orders.ts   imported by index.ts, deployed with it
├── package.json    npm dependencies, installed at build time
└── .zavu/          created by the CLI; do not edit

• Several files are fine. deploy starts at index.ts and uploads every file it reaches through relative imports. Split the prompt, the tools and the business logic however you like.
• Only what is imported ships. A file nothing imports stays on your machine, so tests and scratch work are not deployed.
• npm packages are not files. Declare them under dependencies in package.json; they are installed and bundled for you.
• Limits: 200 files, 900,000 bytes total. Paths above the project root (../shared/x.ts), node_modules/, and names starting with __zavu are refused. To share code across projects, publish it as an npm package.
• --source <file> names a different entrypoint.

When to use Functions vs the imperative AI Agent API