[](https://eve.dev/)

# The framework for
building agents

$npx eve@latest init my-agent

[Read the docs](https://eve.dev/docs/getting-started)

Like Next.js for web apps, but for agents. Markdown for instructions and skills, TypeScript for tools. Durable by default.

## An agent is a directory

Define instructions and skills in markdown, tools in TypeScript, and deploy. The framework compiles the directory, wires up durable workflows, and connects channels.

An instructions.md file is a complete agent. Describe its role in Markdown, then run eve.

eve uses a default model. Add agent.ts when you want to choose a model or configure the runtime.

Leverages

[AI Gateway](/ai-gateway)

Skills are Markdown playbooks loaded when they are relevant. The agent gets focused guidance without carrying it in every prompt.

Add a TypeScript file to tools/ and the model can call it. The filename becomes the tool name. No registration required.

Every agent includes an isolated sandbox and file tools. Add sandbox/sandbox.ts to choose a backend or customize its setup.

Leverages

[Vercel Sandbox](/sandbox)

Add channel files to use the same agent in Slack, Discord, Teams, or the web.

Leverages

[Vercel Connect](/connect) [Chat SDK](/chat)

Connections handle authentication for services such as GitHub, Stripe, and Linear. Tools can call them without managing tokens.

Leverages

[Vercel Connect](/connect)

Add subagents for specialized work. The main agent delegates tasks and combines the results.

Schedules run agents automatically for jobs such as daily reports and weekly digests. Work continues durably without an active session.

Leverages

[Vercel Workflows](/workflows)

agent/

instructions.md

```
# Identity

You are an expert weather assistant.
You can fetch the weather for any
city in the world.
```

Start with instructions.md

```
# Identity

Choose your model in agent.ts

Leverages

[AI Gateway](/ai-gateway)

```
import { defineAgent } from "eve";

export default defineAgent({
  model: "xai/grok-4.5",
});
```

Add reusable skills/

```
---
description: Research unfamiliar topics
---

When the task is novel or ambiguous,
gather evidence first, then answer.
```

Define tools/ in TypeScript

```
import { defineTool } from "eve/tools";
import { z } from "zod";

export default defineTool({
  description: "Get the weather for a city",
  inputSchema: z.object({
    cityName: z.string(),
  }),
  async execute(input) {
    const res = await fetch(
      `${process.env.WEATHER_API_URL}/current?city=${input.cityName}`
    );
    const data = await res.json();
    return data.current_condition[0];
  },
});
```

Customize the sandbox/

Leverages

[Vercel Sandbox](/sandbox)

```
import {
  defineSandbox,
  vercelSandboxBackend,
} from "eve/sandbox";

export default defineSandbox({
  backend: vercelSandboxBackend({
    runtime: "node24",
  }),
});
```

Connect channels/

Leverages

```
import { connectSlackCredentials } from "@vercel/connect/eve";
import { slackChannel } from "eve/channels/slack";

export default slackChannel({
  credentials: connectSlackCredentials("slack/my-agent"),
});
```

Create connections/

Leverages

[Vercel Connect](/connect)

```
import { connect } from "@vercel/connect/eve";
import { defineMcpClientConnection } from "eve/connections";

export default defineMcpClientConnection({
  url: "https://mcp.linear.app/sse",
  description: "Linear workspace: issues, projects, cycles, and comments.",
  auth: connect("linear"),
});
```

Delegate to subagents/

export default defineAgent({
  description: "Investigate questions",
  model: "openai/gpt-5.4",
});
```

Run schedules/

Leverages

```
---
cron: "0 8 * * *"
---

Send the user a daily weather
digest for their saved cities.
```

## Leverages all Vercel AI primitives

AI Gateway for model calls, Sandboxes, Workflows, and Connect. All critical agent infrastructure works out of the box, no gluing point solutions together.

RuntimeDurable execution, state persistence, event streaming.

[

Vercel WorkflowsCheckpointed steps, park between messages, resume on delivery.

](/workflows)

[

AI GatewayModel calls, streaming.

](/ai-gateway) [

Vercel SandboxIsolated execution.

](/sandbox) [

Vercel ConnectMCP/HTTP endpoints.

](/connect)

Tools and subagentsFunctions, child agents.

ChannelWhere your agent gets surfaced.

[

Chat SDK

SlackGoogle ChatDiscordMicrosoft TeamsWeb ChatWhatsAppAPITwilioCronLinear

](/chat)

## Everything you need for production agents

Enterprise governance, observability, and sandboxed compute come standard. Focus on building, not infrastructure.

- resolve\_identity

    customer\_summary

    search\_context

    execute\_sql

    data\_integrity

    [Durable execution(opens in new tab)](https://eve.dev/docs/concepts/execution-model-and-durability)Workflows survive crashes and restarts. Every step is checkpointed. Agents park when waiting, resume on the next message.

- agent\_JKdWWzHCUoj7gKBqnTRunning4xCPU 8GB

    agent\_yN8upsY7Kgh4DFTiYHyxStopped2xCPU 4GB

    agent\_zDiBp4lFo7hYyv35y06DfRunning2xCPU 8GB

    agent\_fORPyGtLtxG4VdlDtRnprStopping4xCPU 8GB

    agent\_T0BqwUg0ierWhp6cD2TStopped2xCPU 8GB

    agent\_SxbECNxlPDoaAFSOZzwRunning4xCPU 8GB

    agent\_Qm2vTnLx8RpKdWe3BfHaRunning2xCPU 4GB

    agent\_Hs9YpZcV4NtLmQr7XaPoStopping4xCPU 8GB

    agent\_Lk4BdWnGpY6xTfRoZ2UyStopped2xCPU 8GB

    agent\_Vc8MqJhDsK1uNbWp5YeoRunning4xCPU 8GB

    [Sandboxed compute(opens in new tab)](https://eve.dev/docs/sandbox)Agents spin up isolated VMs on demand. File system access, bash execution, and code runs, all completely isolated.

- GitHubSlackDiscordMessenger

    Microsoft TeamsGoogle ChatWhatsApp

    LinearTwilioCronWeb ChatAPI

    [Multi-channel delivery(opens in new tab)](https://eve.dev/docs/channels/overview)One agent codebase deploys to web chat, Slack, API, cron, CLI, and custom apps.

- [Human-in-the-loop(opens in new tab)](https://eve.dev/docs/human-in-the-loop)Tools that need confirmation trigger approval gates. Sessions park until resolved, then resume seamlessly.

- [Subagents(opens in new tab)](https://eve.dev/docs/subagents)Delegate specialized work to child agents with their own prompts, tools, and sandbox.

- [Evaluations(opens in new tab)](https://eve.dev/docs/evals/overview)Define test suites with scoring rubrics. Run evals on every deployment and on a schedule.

## Build your first agent

[Get started](https://eve.dev/docs/getting-started)