AI Resources: Use Our Docs With Your AI Tools

Machine-readable documentation

Every page in our documentation is available in multiple formats designed for AI tools and LLMs.

llms.txt

Our llms.txt file follows the llms.txt standard — a markdown index of our most important documentation, designed for AI crawlers and LLMs.

https://theaiuniversity.com/llms.txt

This gives AI systems a curated map of our docs with descriptions, similar to how robots.txt works for search engines.

llms-full.txt

For complete context, we provide llms-full.txt — the entire documentation concatenated into a single file. Feed this to an LLM when you want full knowledge of our platform.

https://theaiuniversity.com/llms-full.txt

Markdown endpoints

Every documentation page is available as clean markdown by appending .md to the URL:

https://theaiuniversity.com/docs/getting-started/quickstart.md

This returns the raw markdown content with proper Content-Type: text/markdown headers. Use this to feed individual pages to AI tools.

Copy as Markdown

Every documentation page has a "Copy as Markdown" button in the header. Click it to copy the full page content to your clipboard in markdown format — ready to paste into any AI tool.

AI prompts for your coding tools

Copy these prompts into your AI coding tool to get better assistance when building agents using our patterns.

Claude Code (CLAUDE.md)

Add this to your project's CLAUDE.md file:

## AI Agent Development

This project follows The AI University's agent architecture patterns:
- Agents are autonomous programs that loop: observe → think → act
- Each agent has a system prompt, tools (MCP), and memory
- Tools are defined with clear descriptions that the LLM reads to decide when to use them
- Use tool allowlists to restrict which tools each agent can access
- Implement guardrails: input validation, rate limiting, human escalation
- Memory: save_memory/load_memory for persistent state across runs
- Architecture: supervisor pattern for multi-agent orchestration

Reference: https://theaiuniversity.com/docs/building-agents/architecture-patterns.md
Reference: https://theaiuniversity.com/docs/tools/overview.md

Cursor Rules (.cursor/rules/)

Create a file at .cursor/rules/ai-agents.mdc:

---
description: AI agent development patterns
globs: ["**/*.ts", "**/*.tsx"]
---

When building AI agents, follow these patterns:
- Define tools with descriptive names and descriptions (the LLM reads descriptions to decide tool usage)
- Use TypeScript for agent code with the Anthropic SDK
- Implement the agent loop: send message → check for tool calls → execute tools → feed results back → repeat
- Always add error handling around tool execution
- Use model routing: Haiku for classification, Sonnet for general tasks, Opus for complex reasoning
- Implement tool allowlists per agent for security
- Add rate limiting and daily budget caps
- Log every tool call and decision for debugging

Docs: https://theaiuniversity.com/llms-full.txt

GitHub Copilot (.github/copilot-instructions.md)

When working on AI agent code:
- Follow the observe-think-act agent loop pattern
- Define MCP tools with name, description, and input_schema
- Tool descriptions should explain WHEN to use the tool, not just what it does
- Implement guardrails: validate inputs, rate limit tool calls, add human escalation
- Use the supervisor pattern for multi-agent orchestration
- Reference: https://theaiuniversity.com/llms.txt

Windsurf Rules

Add to your Windsurf project rules:

For AI agent development:
- Agents loop: prompt LLM → execute tool calls → feed results back → repeat until done
- Tools need clear descriptions for the LLM to know when to use them
- Implement error handling, rate limiting, and budget caps
- Use allowlists to restrict tool access per agent
- For multi-agent systems, use the supervisor pattern with specialized agents

Using our docs with AI assistants

Feed a single page

curl -s https://theaiuniversity.com/docs/building-agents/architecture-patterns.md | pbcopy

Then paste into your AI tool with your question.

Feed all docs at once

curl -s https://theaiuniversity.com/llms-full.txt | pbcopy

This copies our entire documentation to your clipboard. Paste it as context when you need comprehensive knowledge about our platform.

In Claude Code

claude -p "Using the AI University's agent patterns from https://theaiuniversity.com/llms-full.txt, help me build a research agent that searches the web and produces structured reports."

Key takeaways

• Every page is available as markdown via the .md suffix on any URL
• llms.txt provides a curated index for AI crawlers
• llms-full.txt dumps all documentation into one file for full context
• Copy-paste prompts are available for Cursor, Claude Code, Copilot, and Windsurf
• The "Copy as Markdown" button on every page makes it easy to feed content to any AI tool