Skip to main content

Using LLM Files

Integrating Burger API's LLM files into your workflow is straightforward. If you've created your project using the Burger API CLI, you can find these files in the ecosystem/.llm-context/ directory.

This guide covers how to use them with the most popular AI-powered code editors.

For Cursor IDE​

Cursor is built specifically for AI-assisted coding and makes it very easy to reference external files.

1. Attach Files with @ Mentions​

In the Chat or Composer window, simply type @ followed by the name of the LLM file you want to use.

Example Prompt:

@ecosystem/.llm-context/llms-small.txt Generate a new route file for user authentication with login and register endpoints.

2. Best Practices​

  • Use llms-small.txt for most tasks to keep the AI response fast and focused.
  • Use llms-full.txt when you need the AI to understand complex configurations or advanced features.

For GitHub Copilot​

GitHub Copilot can reference files using several methods depending on your IDE (VS Code, IntelliJ, etc.).

1. Reference in Chat​

Open the GitHub Copilot Chat and reference the file path directly.

Example Prompt:

Using the documentation in ./ecosystem/.llm-context/llms-small.txt, explain how to implement a custom middleware for logging.

2. Open Files in Tabs​

Copilot often uses open files as context. Keeping the relevant llm-*.txt file open in a background tab can help Copilot provide more accurate suggestions in other files.

For Claude Code​

Claude Code (Anthropic's CLI tool) is exceptionally powerful at processing documentation context.

1. Using /add​

In your Claude Code session, use the /add command to bring the context files into the conversation.

Example:

/add ecosystem/.llm-context/llms-small.txt Now, using that context, create a route group for 'admin' with a custom authentication middleware.

2. Context Chaining​

Claude's large context window allows you to keep llms-full.txt active for complex architectural discussions without losing track of your project state.

For Other AI Editors​

Whether you are using open-source editors (like PearAI, Continue, or Void) or premium tools like Windsurf:

  • Mentions & Symbols: Most AI editors now use the @ or # symbols to reference files. Try @llms-small.txt first.
  • System Prompts: In tools like Continue, you can add the content of llms-small.txt to your system prompt to always have Burger API context available.
  • Copy/Paste: For specific tasks, copy the relevant section from llms-full.txt directly into your prompt.
  • File Upload: Many modern AI interfaces allow you to upload text files directly as context.
  • Reference URL: If your files are hosted online (e.g., on GitHub), you can provide the raw link to the AI.

Example Prompts​

Here are some effective ways to phrase your requests:

  • Create a Route: "Using the Burger API documentation from @llms-small.txt, create a route file with GET and POST handlers for products."
  • Implement Middleware: "Reference @llms-full.txt to show me how to implement authentication middleware that checks for a Bearer token."
  • Check Validation: "Based on @llms.txt, what validation options are available for request bodies in Burger API?"

Best Practices for AI Prompting​

  1. Be Specific: Tell the AI exactly which file to use and what you want to achieve.
  2. Mind the Tokens: If you're working on a small feature, llms-small.txt is much more efficient than the full version.
  3. Keep Files Updated: When you upgrade your Burger API version, ensure you also update your local copies of the LLM context files.
  4. Combine Context: For the best results, mention both the LLM context file AND the specific file in your project you want to modify.
info

The more context you provide about your specific needs, the better the AI can tailor the Burger API patterns to your project.