LLM Agent Integration

This document describes how to use LimeLink's llms.txt as a data source for various AI coding agents and assistants.

The llms.txt file provides structured documentation that LLM agents can consume to understand LimeLink's API, SDK integration, and dynamic link features.

What is llms.txt?

llms.txt is a standardized format for providing LLM-friendly documentation. It allows AI agents to quickly understand your service's capabilities, API endpoints, and integration patterns.

LimeLink's llms.txt endpoint:

https://limelink.org/llms.txt

Use Cases

• AI-assisted development with LimeLink APIs
• Automated code generation for dynamic link integration
• Context-aware suggestions for SDK implementation
• Documentation lookup during development

Agent Integration Examples

Claude Code

Claude Code can fetch and use llms.txt as context for understanding LimeLink's API.

Using WebFetch tool:

# In Claude Code, use the built-in WebFetch capability
claude "Fetch https://limelink.org/llms.txt and explain the API structure"

With .claude/commands (Custom Slash Command):

Create .claude/commands/limelink.md:

Fetch the LimeLink documentation from https://limelink.org/llms.txt and use it as context for the following task:

$ARGUMENTS

Then use:

/limelink create a dynamic link with iOS deep linking

With MCP Server:

{
  "mcpServers": {
    "fetch": {
      "command": "npx",
      "args": ["-y", "@anthropics/mcp-fetch"]
    }
  }
}

Cursor

Cursor can use llms.txt through its documentation context feature.

Method 1: Add to .cursorrules

# Project Context

For LimeLink integration, refer to: https://limelink.org/llms.txt

When implementing dynamic links:
- Use the API documentation from llms.txt
- Follow the SDK integration patterns
- Apply platform-specific configurations for iOS/Android

Method 2: Use @Web in Chat

@Web https://limelink.org/llms.txt

How do I create a dynamic link with custom preview?

Method 3: Add to Docs

1. Open Cursor Settings → Features → Docs
2. Add new documentation:
   • Name: LimeLink
   • URL: https://limelink.org/llms.txt
3. Use @LimeLink in chat to reference the documentation

OpenAI Codex CLI

Codex can consume llms.txt for context-aware code generation.

Direct fetch in prompt:

codex "Based on https://limelink.org/llms.txt, generate a Node.js function to create a dynamic link"

With context file:

# Download llms.txt locally
curl -o limelink-docs.txt https://limelink.org/llms.txt

# Use as context
codex --context limelink-docs.txt "Create an iOS deep link configuration"

In codex.config.json:

{
  "context": {
    "urls": ["https://limelink.org/llms.txt"]
  }
}

Google Gemini CLI

Gemini CLI can use llms.txt for understanding LimeLink's API structure.

Direct usage:

gemini "Read https://limelink.org/llms.txt and show me how to integrate Android App Links"

With grounding:

gemini --grounding-source "https://limelink.org/llms.txt" \
  "Generate a curl command to create a marketing campaign link"

Using context file:

# Fetch and save
curl -s https://limelink.org/llms.txt > /tmp/limelink.txt

# Use with Gemini
gemini -f /tmp/limelink.txt "What UTM parameters are supported?"

Windsurf (Codeium)

Windsurf can leverage llms.txt for intelligent code suggestions.

Method 1: Cascade with URL context

@URL https://limelink.org/llms.txt

Implement a share function that creates a dynamic link with social preview

Method 2: Add to workspace rules

Create .windsurfrules:

context:
  documentation:
    - name: LimeLink API
      url: https://limelink.org/llms.txt
      description: Dynamic link creation and SDK integration

rules:
  - When implementing deep links, always check LimeLink documentation
  - Use platform-specific options for iOS and Android

Method 3: Memory feature

Remember that LimeLink API documentation is at https://limelink.org/llms.txt.
Use it when I ask about dynamic links or deep linking.

Best Practices

1. Cache the Documentation

For faster responses, cache llms.txt locally:

# Daily update script
curl -s https://limelink.org/llms.txt > ~/.config/limelink-docs.txt

2. Combine with Project Context

Include project-specific information alongside llms.txt:

# Context

- LimeLink Docs: https://limelink.org/llms.txt
- Project API Key: (stored in .env as LIMELINK_API_KEY)
- Target platforms: iOS, Android

3. Validate Generated Code

Always verify AI-generated code against the official API documentation:

• Check required parameters
• Validate response handling
• Test error scenarios

llms.txt Structure

LimeLink's llms.txt includes:

Notes

• llms.txt is updated alongside API changes
• AI agents should handle network errors gracefully
• For production use, implement proper error handling
• Rate limiting may apply to frequent requests