MCP Server

The intellinode command starts an MCP server that exposes the Gen functions and multi-provider chat as tools. A coding assistant connected to it can route work to OpenAI, Anthropic, Gemini, Mistral, Cohere, NVIDIA, OpenRouter, DeepSeek, Groq, xAI, Together or a local Ollama model.

Start the server

npx -y intellinode mcp                       # stdio, for Claude Code, Cursor, VS Code
npx -y intellinode mcp --http --port 3210    # Streamable HTTP at http://127.0.0.1:3210/mcp
npx -y intellinode help

HTTP options: --port (default 3210), --host (default 127.0.0.1), --path (default /mcp), --origin <url> (extra allowed browser origin, repeatable), --debug.

Keys

Keys come from the environment or a .env file in the directory where the server starts:

Variable	Provider
OPENAI_API_KEY	OpenAI (chat tools and image generation)
ANTHROPIC_API_KEY	Anthropic
GEMINI_API_KEY	Google Gemini
MISTRAL_API_KEY	Mistral
COHERE_API_KEY	Cohere
NVIDIA_API_KEY	NVIDIA
OPENROUTER_API_KEY, OPENROUTER_MODEL	OpenRouter (default model openai/gpt-5.5)
DEEPSEEK_API_KEY, DEEPSEEK_MODEL	DeepSeek (default model deepseek-chat)
GROQ_API_KEY and GROQ_MODEL	Groq
XAI_API_KEY and XAI_MODEL	xAI
TOGETHER_API_KEY and TOGETHER_MODEL	Together
OLLAMA_MODEL	A local Ollama model, no key needed
STABILITY_API_KEY	Stability AI images

The default provider is the first configured one in that order; every tool accepts a provider argument to pick another. A missing key returns a tool error naming the variable to set, so the assistant can tell you what to configure.

Install in your assistant

Claude Code:

claude mcp add intellinode -e OPENAI_API_KEY=sk-... -e ANTHROPIC_API_KEY=sk-ant-... -- npx -y intellinode mcp

Or a project .mcp.json (Claude Code) / .cursor/mcp.json (Cursor):

{
  "mcpServers": {
    "intellinode": {
      "command": "npx",
      "args": ["-y", "intellinode", "mcp"],
      "env": { "OPENAI_API_KEY": "sk-...", "ANTHROPIC_API_KEY": "sk-ant-..." }
    }
  }
}

VS Code (.vscode/mcp.json):

{
  "servers": {
    "intellinode": {
      "type": "stdio",
      "command": "npx",
      "args": ["-y", "intellinode", "mcp"],
      "env": { "OPENAI_API_KEY": "sk-..." }
    }
  }
}

For the HTTP transport start npx -y intellinode mcp --http and register { "type": "http", "url": "http://127.0.0.1:3210/mcp" }.

Tools

Tool	Arguments	Returns
ask_model	prompt, provider?, model?, system?	text
consensus	prompt, providers?	the answer of every configured provider side by side
review_code	code, language?, provider?	{ summary, score, issues }
fix_code	code, problem?, language?, provider?	{ code, explanation, changes }
generate_unit_tests	code, framework?, modulePath?, provider?	test file
generate_component	description, framework?, language?, styling?, provider?	code
generate_form	description, framework?, action?, provider?	code
generate_sql	description, dialect?, schema?, provider?	SQL
generate_openapi_spec	input, title?, basePath?, provider?	OpenAPI document
generate_design_tokens	brand, brandColor?, provider?	tokens, CSS variables, Tailwind theme
generate_regex	description, language?, provider?	{ pattern, flags, explanation, matches, nonMatches, verified }
generate_mock_data	schema, count?, provider?	{ records, count }
generate_seo_meta	page, url?, siteName?, provider?	meta fields and rendered HTML
generate_image	prompt, provider? (openai or stability), size?	PNG image block
list_providers	none	configured providers, defaults and the variables still unset

Build your own server

MCPServer serves your own tools over stdio or Streamable HTTP, with no extra dependency:

const { MCPServer } = require('intellinode');

const server = new MCPServer({
  name: 'my-tools',
  version: '1.0.0',
  instructions: 'Optional guidance for the model.',
  tools: [{
    name: 'add',
    description: 'Add two numbers',
    inputSchema: { type: 'object', properties: { a: { type: 'number' }, b: { type: 'number' } }, required: ['a', 'b'] },
    handler: async ({ a, b }) => ({ sum: a + b }),
  }],
});

server.startStdio();                                       // stdio transport
const { url } = await server.startHttp({ port: 3210 });    // or Streamable HTTP, url: http://127.0.0.1:3210/mcp
await server.stop();

A handler may return a string, an object (sent as structuredContent plus a JSON text block), a content block or a full { content, structuredContent, isError } result; a thrown error becomes isError: true with the message. startHttp accepts host, port, path, allowedOrigins and maxBodyBytes.

Registry

The server is listed in the MCP registry as io.github.intelligentnode/intellinode, so clients that browse the registry can install it directly.