> For the complete documentation index, see [llms.txt](https://docs.graphlit.dev/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.graphlit.dev/sdk-setup/for-typescript-developers.md).

# TypeScript

Build AI applications with TypeScript or JavaScript using the Graphlit SDK.

{% hint style="info" %}
**New to Graphlit?** Complete the [Quickstart tutorial](/getting-started/quickstart.md) for a hands-on introduction.
{% endhint %}

***

## Installation

Install the Graphlit client with npm or yarn:

{% tabs %}
{% tab title="npm" %}

```bash
npm install graphlit-client
```

{% endtab %}

{% tab title="yarn" %}

```bash
yarn add graphlit-client
```

{% endtab %}

{% tab title="pnpm" %}

```bash
pnpm add graphlit-client
```

{% endtab %}
{% endtabs %}

**Requirements:**

* Node.js 20 or higher (Docker: use `node:20-slim` or `node:22-slim`)
* Graphlit account with [API credentials](/account-setup-one-time/create-project.md)

***

## Quick Start

{% tabs %}
{% tab title="TypeScript" %}

```typescript
import { Graphlit } from 'graphlit-client';

async function main() {
  const graphlit = new Graphlit();

  const response = await graphlit.ingestText(
    'Our AI agent needs persistent memory across sessions...',
    'Product Requirements',
  );

  console.log(`✅ Memory created: ${response.ingestText.id}`);
}

main().catch((error) => {
  console.error(error);
  process.exit(1);
});
```

{% endtab %}

{% tab title="JavaScript (ESM)" %}

```javascript
import { Graphlit } from 'graphlit-client';

async function main() {
  const graphlit = new Graphlit();

  const response = await graphlit.ingestText(
    'Our AI agent needs persistent memory across sessions...',
    'Product Requirements'
  );

  console.log(`✅ Memory created: ${response.ingestText.id}`);
}

main().catch((error) => {
  console.error(error);
  process.exit(1);
});
```

{% endtab %}

{% tab title="JavaScript (CommonJS)" %}

```javascript
require('dotenv/config');
const { Graphlit } = require('graphlit-client');

async function main() {
  const graphlit = new Graphlit();

  const response = await graphlit.ingestText(
    'Our AI agent needs persistent memory across sessions...',
    'Product Requirements'
  );

  console.log(`✅ Memory created: ${response.ingestText.id}`);
}

main().catch((error) => {
  console.error(error);
  process.exit(1);
});
```

{% endtab %}
{% endtabs %}

{% hint style="success" %}
**That's it!** You now have semantic memory for your AI application.
{% endhint %}

***

## Configuration Options

### Environment Variables (Recommended)

```typescript
import { Graphlit } from 'graphlit-client';

const graphlit = new Graphlit();
```

{% hint style="warning" %}
**Security:** Never commit credentials to git. Use environment variables or secrets management in production.
{% endhint %}

### Alternative: Explicit Configuration

Only use if you need to override environment variables:

```typescript
import { Graphlit } from 'graphlit-client';

const graphlit = new Graphlit({
  organizationId: process.env.GRAPHLIT_ORGANIZATION_ID,
  environmentId: process.env.GRAPHLIT_ENVIRONMENT_ID,
  jwtSecret: process.env.GRAPHLIT_JWT_SECRET,
});
```

***

## Common Patterns

### Ingest Content

```typescript
// From URL
const pdf = await graphlit.ingestUri(
  'https://example.com/document.pdf',
  'Product Brief',
  undefined, // id
  undefined, // identifier
  true,      // isSynchronous
);

console.log(`📄 PDF ready: ${pdf.ingestUri.id}`);

// From text
const notes = await graphlit.ingestText(
  'Discussion about Q4 planning...',
  'Meeting Notes',
);

console.log(`📝 Notes ready: ${notes.ingestText.id}`);
```

### Search Memory

```typescript
const response = await graphlit.queryContents({
  search: 'quarterly planning'
});

for (const content of response.contents?.results ?? []) {
  console.log(`📄 ${content.name}`);
}
```

### Chat with Context

```typescript
// Create conversation
const conversation = await graphlit.createConversation({
  name: 'AI Assistant',
});

// Ask questions
const answer = await graphlit.promptConversation(
  'What did we discuss about Q4 planning?',
  conversation.createConversation.id,
);

console.log(answer.promptConversation.message?.message);
```

***

## Next Steps

**Quickstarts:**

* [Quickstart: Your First Agent](/getting-started/quickstart.md) - Build a streaming agent in 7 minutes
* [AI Agents](/tutorials/ai-agents.md) - Create agents with persistent memory
* [MCP Integration](/mcp-integration/mcp-integration.md) - Connect to your IDE

**Examples:**

* [Next.js Applications](https://github.com/graphlit/graphlit-samples/tree/main/nextjs) - Full-stack chat apps
* [MCP Server](https://github.com/graphlit/graphlit-mcp-server) - Production MCP implementation

**Resources:**

* [TypeScript SDK on GitHub](https://github.com/graphlit/graphlit-client-typescript)
* [Use Case Library](/api-guides/use-cases.md) - 100+ code examples
* [Ask Graphlit](/resources/ask-graphlit.md) - AI code assistant
* [Join Discord](https://discord.gg/ygFmfjy3Qx) - Get help from the community