> ## Documentation Index
> Fetch the complete documentation index at: https://docs.usebruno.com/llms.txt
> Use this file to discover all available pages before exploring further.

# External Libraries

Bruno allows you to load any npm module or package for use in your scripting workflows.

<Warning>
  Always turn on [Developer Mode](/get-started/configure/javascript-sandbox) when using external libraries in Bruno.
</Warning>

## Prerequisites

Before you begin, ensure that you have the following:

1. Node.js installed on your system.
2. Bruno installed.

## Project structure

The folder structure for this example will look like this:

<Tree>
  <Tree.Folder name="bruno-collection" defaultOpen>
    <Tree.Folder name="external-lib" defaultOpen>
      <Tree.File name="echo-api.bru" />

      <Tree.File name="auth.bru" />
    </Tree.Folder>

    <Tree.File name="data.csv" />

    <Tree.File name="package.json" />

    <Tree.File name="bruno.json" />

    <Tree.File name=".gitignore" />

    <Tree.Folder name="node_modules" />
  </Tree.Folder>
</Tree>

## Configuration

You need a `package.json` file at the root of your Bruno collection.

1. Open your collection folder in **Bruno** or in an editor such as **VS Code**.
2. Open the integrated terminal (in Bruno or your editor) and go to the collection root:

```bash theme={null}
cd path/to/your/bruno-collection
```

3. Initialize `package.json`:

```bash theme={null}
npm init -y
```

4. Install the package (run this in the same terminal, from the collection root):

```bash theme={null}
npm install csv-parse
```

5. Confirm **`csv-parse`** appears under `dependencies` in `package.json`, for example:

```json theme={null}
"dependencies": {
  "csv-parse": "^6.2.1"
}
```

Exact versions may differ; what matters is that the dependency is listed after a successful install.

6. In Bruno, enable **[Developer Mode](/get-started/configure/javascript-sandbox)** for the collection so `require()` can load packages from `node_modules`.

## Example: Parse CSV with csv-parse

This example uses [csv-parse <strong><sup>↗</sup></strong>](https://www.npmjs.com/package/csv-parse) to turn CSV text into JavaScript objects, then sends the first row as JSON on the request. See the [npm package page <strong><sup>↗</sup></strong>](https://www.npmjs.com/package/csv-parse) for options, streaming APIs, and more.

Use a **Pre Request** script with **Developer Mode** on:

```javascript theme={null}
const { parse } = require('csv-parse/sync');

const csvData = `name,email
Smith,smith@example.com
John,john@example.com`;

const records = parse(csvData, {
  columns: true,
  skip_empty_lines: true
});

req.setHeader('Content-Type', 'application/json');
req.setBody(records[0]);
```

<img src="https://mintcdn.com/bruno-a6972042/seGko4JB7JPOWh_n/images/screenshots/script/external-lib.webp?fit=max&auto=format&n=seGko4JB7JPOWh_n&q=85&s=6f88af94e1eac0582bc561155f653a7d" alt="Bruno Pre Request script using an external library (csv-parse)" width="2618" height="1296" data-path="images/screenshots/script/external-lib.webp" />

`records[0]` is the first data row as an object (for example `{ name: 'Smith', email: 'smith@example.com' }`) because `columns: true` uses the first line as headers.

### Reading CSV from a file

You can also pass CSV from a file in your collection (path is relative to the collection root):

```javascript theme={null}
const fs = require('fs');
const path = require('path');
const { parse } = require('csv-parse/sync');

const csvData = fs.readFileSync(path.join(__dirname, 'data.csv'), 'utf8');
const records = parse(csvData, {
  columns: true,
  skip_empty_lines: true
});

req.setHeader('Content-Type', 'application/json');
req.setBody(records[0]);
```

Adjust the path to match where you store your `.csv` file.