AI Coding Tools

Make your AI assistant understand the mindzieAPI instantly

Modern AI coding tools like Claude Code, Cursor, Windsurf, and GitHub Copilot can read documentation directly from URLs. We provide LLM-optimized documentation files that give your AI coding assistant complete knowledge of the mindzieAPI.

Quick Reference

Copy these URLs to configure your AI coding tools:

Resource URL Best For
Full Documentation https://docs.mindziestudio.com/llms-full.txt Complete API understanding (~120K tokens)
Documentation Index https://docs.mindziestudio.com/llms.txt Quick reference with links to pages

Claude Code

Claude Code can access documentation using WebFetch or custom instructions.

In a Claude Code session, fetch the complete documentation:

Please read the mindzieAPI documentation from https://docs.mindziestudio.com/llms-full.txt

Claude will fetch and read the entire API documentation. You can then ask questions about endpoints, authentication, request/response formats, and get code examples.

Method 2: Add to Project Instructions

For persistent access across sessions, add to your project's CLAUDE.md or .claude/settings.json:

## mindzieAPI Reference

When working with the mindzieAPI, fetch documentation from:
https://docs.mindziestudio.com/llms-full.txt

This contains complete API documentation including:
- Authentication (Bearer tokens, API keys)
- All endpoints (tenants, users, projects, datasets, blocks, dashboards)
- Request/response formats
- Code examples in multiple languages

Example Prompts

Once Claude has the documentation, you can ask:

  • "How do I authenticate with the mindzieAPI?"
  • "Write Python code to create a new dataset"
  • "What's the endpoint for executing a block?"
  • "Show me how to upload a CSV file to a project"

Cursor IDE

Cursor can index documentation using the @docs feature for instant access during coding.

Setup Steps

  1. Open Cursor Settings (Cmd/Ctrl + ,)
  2. Navigate to Features > Docs
  3. Click Add new doc
  4. Enter URL: https://docs.mindziestudio.com/llms-full.txt
  5. Name it: mindzieAPI

Usage

Reference the documentation in Cursor chat:

@mindzieAPI How do I authenticate API requests?

@mindzieAPI Write a function to list all projects in a tenant

Windsurf

Windsurf supports external documentation sources for AI-assisted coding.

Setup

  1. Open Windsurf settings
  2. Navigate to the knowledge base or documentation section
  3. Add https://docs.mindziestudio.com/llms-full.txt as an external source

Usage

When coding, Windsurf will automatically reference the mindzieAPI documentation to provide accurate suggestions and completions.

GitHub Copilot

While Copilot doesn't directly fetch URLs, you can provide context through project files.

Option 1: Include in Project

Create a docs/mindzieAPI.md file in your project with the API reference. Copilot will use this as context when you're working in that project.

Option 2: Copilot Chat

In GitHub Copilot Chat, paste key sections of the documentation or reference the URL:

Using the mindzieAPI documented at https://docs.mindziestudio.com/llms-full.txt,
write a Python class to manage datasets.

Cody (Sourcegraph)

Cody can index external documentation for context-aware assistance.

Setup

  1. Open Cody settings
  2. Add https://docs.mindziestudio.com/llms-full.txt to your context sources
  3. The documentation will be available across your coding sessions

Generic LLM Usage

For any LLM interface (ChatGPT, Claude web, etc.), you can:

  1. Fetch the index first: Visit https://docs.mindziestudio.com/llms.txt to see the documentation structure
  2. Get complete docs: Copy the content from https://docs.mindziestudio.com/llms-full.txt into your LLM context
  3. Ask questions: The LLM now understands the entire mindzieAPI

What's Included

The LLM documentation covers all mindzieAPI capabilities:

Category Coverage
Authentication API keys (Global and Tenant), Bearer tokens, scopes, security best practices
Tenants Multi-tenant management, creation, updates, deletion with safeguards
Users Global operations, tenant-scoped operations, roles and permissions
Projects CRUD operations, caching, user access, import/export (.mpz files)
Datasets Creation, CSV/XES import, updates, column mapping, file formats
Blocks Analysis blocks, execution, results retrieval, block types
Dashboards Management, panel configuration, sharing and public URLs
Enrichments Pipelines, Python notebook integration, execution
Actions Named action execution, ping endpoints, execution history
Execution Async job management, queue operations, status tracking

Context Window Considerations

Different AI models have different context limits. Here's how our documentation files fit:

Model Context Limit llms-full.txt Recommendation
Claude Opus 4 200K tokens Fits (~120K) Use full documentation
Claude Sonnet 200K tokens Fits (~120K) Use full documentation
GPT-4 Turbo 128K tokens Tight fit Use full documentation
GPT-4o 128K tokens Tight fit Use full documentation
Claude Haiku 200K tokens Fits (~120K) Use full documentation
Gemini Pro 128K tokens Tight fit May need index + specific pages
GPT-3.5 16K tokens Too large Use index, fetch specific pages

For Smaller Context Windows

If your model has limited context:

  1. Use llms.txt (the index) to understand the API structure
  2. Identify which sections you need
  3. Fetch individual markdown files from /docs-master/mindzieAPI/{category}/{page}/page.md

File Formats

URL Format Size Tokens
/llms.txt Markdown (index) ~6 KB ~1.5K
/llms-full.txt Markdown (complete) ~470 KB ~120K
/docs-master/.../*.md Markdown (individual pages) 2-15 KB each ~500-4K each

Keeping Documentation Current

The LLM documentation is regenerated whenever the API documentation is updated. The files include a timestamp showing when they were last generated.

For the most current documentation, your AI tool should fetch fresh copies rather than caching indefinitely.

MCP Server Integration

For advanced AI tool integration, the mindzieAPI provides an MCP (Model Context Protocol) server that enables AI assistants to interact with mindzieStudio programmatically.

The MCP server includes tools like mindzie_list_block_types with a unified category parameter that returns all block types (filters, calculators, and enrichments) with complete metadata in a single call.

View MCP Server Documentation

Next Steps

Ready to start coding with AI assistance? Here are some things to try:

  1. Point your AI coding tool to https://docs.mindziestudio.com/llms-full.txt
  2. Ask: "How do I authenticate with the mindzieAPI?"
  3. Request: "Write Python code to list all projects"
  4. Build: Complete integrations with AI-assisted code generation

For human-readable documentation, see the Quick Start Guide or explore the full API Reference.