close
Skip to main content

Using Juno with AI

Juno provides several ways to integrate with AI tools, from documentation context files to a full MCP server for managing your project through natural language.

  • LLMs.txt: help AI tools better understand the platform
  • MCP Server: interact with Juno directly from your AI tool

LLMs.txt​

An LLMs.txt file is a plain text file that provides instructions or metadata for large language models (LLMs). It often specifies how LLMs should process or interact with content. It's similar to a robots.txt or sitemap.xml file, but tailored for AI models.

Available routes​

We provide several llms.txt routes.

  • llms.txt: Table of contents with links to individual Markdown docs
  • llms-full.txt: Entire documentation in a single Markdown file

Most AI tools work best with one of these formats. Some tools (like Cursor) can benefit from indexing both. Use the combination that works best with your workflow.

How to use it​

Here are some examples of how the llms.txt files can be used with AI tools.

note

🙏 Help us improve! If you use a tool that supports LLMs.txt files, open a pull request to add your example to this page.

Cursor​

You can use custom documentation in Cursor's context using the @Docs feature.

Setup​

To add custom documentation, type @Docs and select Add new doc, or go to Cursor Settings > Features > Docs.

Add both llms.txt URLs:

NameEntry point
Junollms.txt
Juno (full)llms-full.txt

Cursor will index all subpages and use both files to improve context and coverage.

Usage​

Type @Docs in chat to view available docs and select the Juno entries to begin using the references.

important

You must reference the Juno docs in chat using @Docs - Cursor won't use them otherwise.

For example, you could start a conversation with: I want to create a web shop app with Next.js using @Juno and @Juno (full).

Resources​

Read more in the Cursor documentation.

MCP Server​

A Model Context Protocol (MCP) server for Juno is available, built and maintained by a Juno community contributor. It allows MCP-compatible AI tools - such as Claude, Cursor, or VS Code Copilot - to interact with Juno directly through natural language.

It provides tools to manage identity, config, hosting, serverless functions, and changes - with built-in access to Juno's documentation. Long-running operations like deploys and publishes support real-time progress streaming.

note

This is a community-maintained project. It is not part of the core, but we absolutely 💙 it!

Setup​

Follow these steps to connect the MCP server to your AI tool of choice.

1. Add to your MCP client​

No install needed - npx handles downloading automatically.

{
  "mcpServers": {
    "junobuild": {
      "command": "npx",
      "args": ["-y", "junobuild-mcp-server"]
    }
  }
}

2. Authenticate the Juno CLI​

The server wraps the Juno CLI, which must be installed and authenticated:

npm i -g @junobuild/cli
juno login

Once configured, you can interact with Juno directly through your AI tool. Just describe what you want in natural language - for example, "deploy my satellite" or "list my pending changes" - and the AI will call the right tools automatically.

Resources​