AI & LLM Integration

AI & LLM Features

Pushduck documentation provides AI-friendly endpoints that make it easy for large language models (LLMs) and automated tools to access and process our documentation content.

Available Endpoints

Complete Documentation Export

Access all documentation content in a single, structured format:

GET /llms.txt

This endpoint returns all documentation pages in a clean, AI-readable format with:

• Page titles and URLs
• Descriptions and metadata
• Full content with proper formatting
• Structured sections and hierarchies

Example Usage:

curl https://your-domain.com/llms.txt

Individual Page Access

Access any documentation page's raw content by appending .mdx to its URL:

GET /docs/{page-path}.mdx

Examples:

• /docs/quick-start.mdx - Quick start guide content
• /docs/api/client/use-upload-route.mdx - Hook documentation
• /docs/providers/aws-s3.mdx - AWS S3 setup guide

Use Cases

AI Assistant Integration

• Train custom AI models on our documentation
• Create chatbots that can answer questions about Pushduck
• Build intelligent documentation search systems

Development Tools

• Generate code examples and snippets
• Create automated documentation tests
• Build CLI tools that reference our docs

Content Analysis

• Analyze documentation completeness
• Track content changes over time
• Generate documentation metrics

Content Format

The LLM endpoints return content in a structured format:

# Page Title
URL: /docs/page-path

Page description here

# Section Headers
Content with proper markdown formatting...

## Subsections
- Lists and bullet points
- Code blocks with syntax highlighting
- Tables and structured data

Technical Details

• Caching: Content is cached for optimal performance
• Processing: Uses Remark pipeline with MDX and GFM support
• Format: Clean markdown with frontmatter removed
• Encoding: UTF-8 text format
• CORS: Enabled for cross-origin requests

Rate Limiting

These endpoints are designed for programmatic access and don't have aggressive rate limiting. However, please be respectful:

• Cache responses when possible
• Avoid excessive automated requests
• Use appropriate user agents for your tools

Examples

Python Script

import requests

# Get all documentation
response = requests.get('https://your-domain.com/llms.txt')
docs_content = response.text

# Get specific page
page_response = requests.get('https://your-domain.com/docs/quick-start.mdx')
page_content = page_response.text

Node.js/JavaScript

// Fetch all documentation
const allDocs = await fetch("/llms.txt").then((r) => r.text());

// Fetch specific page
const quickStart = await fetch("/docs/quick-start.mdx").then((r) => r.text());

cURL

# Download all docs to file
curl -o pushduck-docs.txt https://your-domain.com/llms.txt

# Get specific page content
curl https://your-domain.com/docs/api/client/use-upload-route.mdx

Integration with Popular AI Tools

OpenAI GPT

Use the /llms.txt endpoint to provide context about Pushduck in your GPT conversations.

Claude/Anthropic

Feed documentation content to Claude for detailed analysis and code generation.

Local LLMs

Download content for training or fine-tuning local language models.

These AI-friendly endpoints make it easy to integrate Pushduck documentation into your development workflow and AI-powered tools!