AI-Assisted Integration

Nilo provides an AI-optimized documentation file (llms.txt) that allows AI coding assistants like Cursor, GitHub Copilot, Claude, or ChatGPT to understand the entire Nilo API and help you generate integration code in any programming language.

What is llms.txt?

The llms.txt file is a condensed, structured version of the entire Nilo Integration API documentation, designed specifically for AI tools to consume. It contains all endpoints, parameters, request/response examples, webhook payloads, and common integration patterns in a single file.

You can find it at:

https://developers.nilo-labs.com/llms.txt

How to Use It

With Cursor

Paste the URL as context when asking Cursor to generate integration code:

@https://developers.nilo-labs.com/llms.txt

I need to sync my SAP product catalog with Nilo.
Generate a C# service that authenticates, then batch upserts
all products from my database every 4 hours.

Cursor will read the full API reference and generate correct, working code in your language of choice.

With Claude or ChatGPT

Copy the contents of the llms.txt file and paste it into the conversation as context, then ask your question:

Here is the Nilo Integration API documentation:
[paste llms.txt content]

How do I set up a webhook endpoint in Java Spring Boot to receive
new orders from Nilo and confirm them automatically?

With Claude Projects

Upload the llms.txt file as a project knowledge file. This creates a persistent assistant that knows the entire Nilo API and can help you with any integration question across multiple conversations.

What You Can Ask

Here are some examples of what AI tools can help you with using the llms.txt file:

• "Generate a full integration service in [your language]" — The AI can generate authentication, CRUD operations, batch syncs, and webhook handlers in any language
• "How do I sync only changed products from my ERP?" — The AI will suggest using batch upsert with delta detection patterns
• "Create a webhook handler for receiving orders" — The AI will generate the correct endpoint implementation based on the webhook payload structure
• "What's the correct order to set up my catalog?" — The AI knows the recommended setup sequence (groupers → brands → categories → products → stock → prices → stores)

Supported Languages

Since llms.txt is language-agnostic, AI tools can generate integration code in any language:

• JavaScript / TypeScript (Node.js)
• Python
• C# / .NET
• Java
• PHP
• Go
• ABAP (SAP)
• Ruby
• Any other language

Tips for Best Results

1. Be specific about your tech stack when asking the AI (e.g., "using .NET 8 with HttpClient" or "using Python requests library")
2. Mention your ERP if relevant (e.g., "I'm using SAP S/4HANA" or "my data comes from Oracle ERP")
3. Ask step by step if the integration is complex — start with authentication, then catalog sync, then order handling
4. Include constraints like "I need to handle rate limits" or "I need retry logic for failed batch operations"