llms.txt
AI-friendly documentation file for code editors like Cursor, Windsurf, and other AI tools
Mesh provides an llms.txt file that contains all documentation and code examples in a standardized, AI-friendly format. You can plug this file into any AI code editor to get instant access to up-to-date Mesh documentation.
Overview
The llms.txt file follows an emerging standard for providing documentation to large language models. It contains:
- Complete API reference for all Mesh packages
- Working code examples with imports
- Common patterns and best practices
- Troubleshooting guides
Because the file uses standardized markdown formatting, AI tools can parse and index it consistently across different platforms.
Use Cases
- Cursor - Add as a doc source for context-aware code completion
- Windsurf - Reference in conversations for accurate Mesh code generation
- ChatGPT/Claude - Upload as context for Mesh-specific questions
- Custom AI pipelines - Parse programmatically for RAG systems
Quick Start
Cursor
- Open Cursor settings
- Navigate to Features > Docs
- Click Add new doc
- Paste:
https://meshjs.dev/llms.txt - Click Add
Once indexed, reference Mesh docs in any chat by typing @Docs and selecting the Mesh documentation.
Windsurf
- In a Windsurf chat, type:
@https://meshjs.dev/llms.txt - The AI now has full Mesh context for your conversation
ChatGPT or Claude
- Download the file: meshjs.dev/llms.txt
- Upload to your conversation as a file attachment
- Ask questions about Mesh SDK
Programmatic Access
import { fetch } from "undici";
async function getMeshDocs(): Promise<string> {
const response = await fetch("https://meshjs.dev/llms.txt");
return response.text();
}File Location
Access the file directly at: meshjs.dev/llms.txt
Benefits Over Traditional Docs
| Traditional Documentation | llms.txt |
|---|---|
| Requires manual searching | AI searches contextually |
| May be outdated in AI training data | Always current |
| Split across multiple pages | Single, complete file |
| Formatted for humans only | Optimized for both humans and AI |
How to Use in Your Projects
Once you add llms.txt to your AI tool, you can ask questions like:
- "How do I build a transaction that sends 5 ADA using Mesh?"
- "Show me how to connect a browser wallet with CIP-30"
- "Generate a minting transaction for an NFT collection"
- "What's the correct order of methods for MeshTxBuilder?"
The AI responds with accurate, up-to-date code based on the latest Mesh documentation.
Troubleshooting
File not indexing in Cursor
- Verify the URL is correct:
https://meshjs.dev/llms.txt - Check your internet connection
- Try removing and re-adding the doc source
- Restart Cursor after adding
AI gives outdated information
- Re-index the documentation in your AI tool's settings
- The
llms.txtfile updates with each Mesh release
File is too large for upload
- Some AI tools have file size limits
- Use the MCP server instead for real-time access: MCP Setup
Related Resources
- MCP Server - Real-time AI integration for VS Code, Cursor, and Claude Desktop
- Claude Code Skills - Deep Mesh knowledge for Claude Code users
- MeshAI Chatbot - Browser-based AI assistant