# How to Use the Netlify Plugin for Serverless Deployments: Complete Setup Guide > Master serverless deployments with the Netlify plugin. Discover automatic function discovery, CLI tooling, and SDK support for seamless builds and testing. - Repository: [OpenAI/plugins](https://github.com/openai/plugins) - Tags: how-to-guide - Published: 2026-07-04 --- **The Netlify plugin provides a complete stack for building, testing, and deploying serverless applications on Netlify through automatic function discovery, CLI tooling, and integrated SDK support.** The `openai/plugins` repository includes a dedicated Netlify plugin that streamlines the creation and deployment of serverless functions. This plugin implements the same architecture that Netlify uses internally, offering a structured approach to building serverless applications with TypeScript and JavaScript. By following the patterns defined in [`plugins/netlify/skills/netlify-functions/SKILL.md`](https://github.com/openai/plugins/blob/main/plugins/netlify/skills/netlify-functions/SKILL.md), you can deploy production-ready endpoints without managing infrastructure. ## Understanding the Netlify Plugin Architecture The plugin organizes serverless deployments into three core components: **Netlify Functions**, **configuration files**, and **CLI tooling**. According to the source code in [`plugins/netlify/skills/netlify-functions/SKILL.md`](https://github.com/openai/plugins/blob/main/plugins/netlify/skills/netlify-functions/SKILL.md), functions are automatically discovered when placed in the `netlify/functions/` directory and exposed at the `/.netlify/functions/` endpoint. ### Function Structure and Discovery Each TypeScript or JavaScript file placed under `netlify/functions/` becomes a standalone serverless endpoint. The plugin expects functions to follow the standard Netlify signature using types from `@netlify/functions`. ### Configuration Management The [`netlify.toml`](https://github.com/openai/plugins/blob/main/netlify.toml) file, documented in [`plugins/netlify/skills/netlify-deploy/SKILL.md`](https://github.com/openai/plugins/blob/main/plugins/netlify/skills/netlify-deploy/SKILL.md), controls build settings, redirect rules, and edge-function routing. Netlify reads this configuration during the build process to generate final deployment assets. ## Setting Up Your First Serverless Function To create a serverless function, add a TypeScript file to the `netlify/functions/` directory. As specified in [`plugins/netlify/skills/netlify-functions/SKILL.md`](https://github.com/openai/plugins/blob/main/plugins/netlify/skills/netlify-functions/SKILL.md), each function receives a `Context` and optional `Config` type from the `@netlify/functions` package. ```typescript // netlify/functions/hello.ts import type { Handler, Context } from "@netlify/functions"; export const handler: Handler = async (event, context: Context) => { return { statusCode: 200, body: JSON.stringify({ message: "👋 Hello from Netlify Functions!" }), }; }; ``` This code automatically becomes available at `/.netlify/functions/hello` once deployed. ## Configuring Routing with netlify.toml The [`netlify.toml`](https://github.com/openai/plugins/blob/main/netlify.toml) file defines how requests route to your functions and static assets. Based on the documentation in [`plugins/netlify/skills/netlify-deploy/SKILL.md`](https://github.com/openai/plugins/blob/main/plugins/netlify/skills/netlify-deploy/SKILL.md), you can configure redirects to map custom paths to function endpoints. ```toml # netlify.toml (project root) [build] functions = "netlify/functions" publish = "dist" [[redirects]] from = "/api/*" to = "/.netlify/functions/:splat" status = 200 ``` This configuration redirects all `/api/*` requests to the corresponding function, hiding the internal `/.netlify/functions/` path from end users. ## Deploying with the Netlify CLI The deployment process uses the Netlify CLI to package your site, upload functions, and generate preview URLs. According to [`plugins/netlify/skills/netlify-cli-and-deploy/SKILL.md`](https://github.com/openai/plugins/blob/main/plugins/netlify/skills/netlify-cli-and-deploy/SKILL.md), the plugin supports both preview and production deployments. First, install the CLI globally: ```bash npm install -g netlify-cli ``` Then deploy a preview build: ```bash netlify deploy --build ``` This command runs the build process, uploads your functions, and returns a unique preview URL. Once verified, deploy to production: ```bash netlify deploy --prod ``` ## Accessing Runtime Primitives and Storage The plugin bundles official Netlify SDKs including `@netlify/functions`, `@netlify/blobs`, and `@netlify/identity`, enabling direct access to platform primitives from within your code. ### Environment Variables Environment variables are automatically injected at build time and accessible via `process.env` without additional tooling. This works out of the box for both local development and deployed functions. ### Blob Storage for Binary Data For large binary payloads such as image uploads, the plugin includes blob storage support documented in [`plugins/netlify/skills/netlify-blobs/SKILL.md`](https://github.com/openai/plugins/blob/main/plugins/netlify/skills/netlify-blobs/SKILL.md). Use the `@netlify/blobs` package to read and write persistent data. ```typescript import { getStore } from "@netlify/blobs"; export const handler = async () => { const store = getStore("my-store"); await store.set("image.png", "binary-data-here"); return { statusCode: 200, body: "Blob stored successfully" }; }; ``` ## Summary - Place serverless functions in `netlify/functions/` to automatically expose them at `/.netlify/functions/` - Configure routing, builds, and redirects in [`netlify.toml`](https://github.com/openai/plugins/blob/main/netlify.toml) as documented in [`plugins/netlify/skills/netlify-deploy/SKILL.md`](https://github.com/openai/plugins/blob/main/plugins/netlify/skills/netlify-deploy/SKILL.md) - Use the Netlify CLI (`netlify deploy --build`) to create preview deployments before pushing to production - Access environment variables through `process.env` and binary storage through `@netlify/blobs` without additional configuration - Reference the skill files in `plugins/netlify/skills/` for detailed implementation patterns and best practices ## Frequently Asked Questions ### How do I install the Netlify plugin from the openai/plugins repository? The Netlify plugin is included in the `openai/plugins` repository under [`plugins/netlify/.codex-plugin/plugin.json`](https://github.com/openai/plugins/blob/main/plugins/netlify/.codex-plugin/plugin.json). You do not need to install the plugin itself separately; instead, install the Netlify CLI and the required SDKs (`npm i @netlify/functions`) in your project to begin using the patterns documented in the skill files. ### What is the correct file structure for Netlify Functions? Place all serverless function files in the `netlify/functions/` directory at your project root. Each file (such as [`hello.ts`](https://github.com/openai/plugins/blob/main/hello.ts)) automatically becomes an endpoint at `/.netlify/functions/`. The build configuration in [`netlify.toml`](https://github.com/openai/plugins/blob/main/netlify.toml) can customize this directory, but `netlify/functions/` is the default convention used by the plugin. ### How do I access environment variables in Netlify Functions? Environment variables are automatically injected at build time and runtime. Access them directly through `process.env.VARIABLE_NAME` within your function code. The plugin handles the injection automatically, so no additional import or configuration is required beyond setting the variables in your Netlify dashboard or local `.env` file. ### Can I test Netlify Functions locally before deploying? Yes. Use the `netlify dev` command to run a local development server that emulates the Netlify Functions environment. This allows you to test endpoints at `http://localhost:8888/.netlify/functions/` before running `netlify deploy --build` to create a preview deployment.