ExisJS MCP Server
The Model Context Protocol (MCP) is an open standard that allows AI coding assistants (like Claude, Cursor, and Google Antigravity) to access external tools and data sources.
ExisJS provides an official MCP Server (@exisjs/mcp) that acts as a bridge between your AI agent and your local ExisJS workspace. It empowers the AI with the ability to dynamically read your framework documentation offline and flawlessly scaffold idiomatic ExisJS boilerplate code (routes, plugins, gateways, configurations).
Without the MCP, AI agents often hallucinate Express or NestJS syntax when asked to write ExisJS code because ExisJS might not be fully present in their training data. By using the ExisJS MCP, the AI reads the exact documentation for your installed version and uses strict generators to scaffold perfect code.
Prerequisites
- Node.js (version >= 20.19) installed on your machine.
- An AI Coding Assistant that supports the Model Context Protocol (e.g., Claude Code, Cursor, Windsurf, or Google Antigravity).
Installation
The easiest way to install the MCP server across all your coding agents is by using the add-mcp CLI:
Add -y to skip the prompt and install to all detected agents. Add -g to install globally across all projects.
Manual Configuration
If you prefer to configure your AI client manually (or are using an unsupported client), you can add the following JSON to your client's MCP settings file:
{ "mcpServers": { "exisjs-mcp": { "command": "npx", "args": ["-y", "@exisjs/mcp@latest"] } }}Claude Code Setup
Run the following command in your terminal:
claude mcp add exisjs-mcp npx @exisjs/mcp@latestCursor Setup
- Open Cursor Settings.
- Navigate to Features > MCP.
- Click + Add New MCP Server.
- Type:
exisjs-mcp - Command:
npx -y @exisjs/mcp@latest
Available Tools
Once the MCP server is installed and running, your AI assistant automatically gains access to the following capabilities:
1. Documentation Readers
ExisJS ships its official markdown documentation directly inside the npm package. The MCP server reads these files locally from your node_modules.
exis_list_local_docs: Recursively scans and lists all available ExisJS documentation topics.exis_read_local_docs: Reads the exact markdown file requested by the AI so it can learn ExisJS best practices.
Because this runs locally, the AI reads the exact documentation that matches the ExisJS version installed in your project, preventing version mismatches!
2. Code Generators
When you ask the AI to build a new feature, it will use these strict scaffolding tools to generate the correct file structures instead of guessing the syntax.
exis_generate_route: Scaffolds an idiomatic ExisJS route controller (Functional or OOP).exis_generate_plugin: Scaffolds a hybrid ExisJS plugin module.exis_generate_gateway: Scaffolds a WebSockets gateway.exis_generate_config: Scaffolds the strictexis.config.tsconfiguration system.
Usage Example
Once configured, simply open your ExisJS project in your AI IDE and ask it a question!
Try asking:
"I want to create a new WebSockets gateway in ExisJS. Can you read the docs on how Gateways work and then scaffold one for me?"
The AI will:
- Use
exis_list_local_docsto find the documentation on Gateways. - Use
exis_read_local_docsto learn the exact syntax for your ExisJS version. - Use
exis_generate_gatewayto safely write the boilerplate into your codebase.