Invompt MCP Server

invompt-mcp

MCP server for Invompt — the AI invoice generator.

Describe what you need, get a professional invoice. Connects your AI assistant to the Invompt API so you can create, format, and share invoices without leaving your editor.

Get started →

What it does

Get your API key

1. Sign up at invompt.com
2. Go to Integrations → Generate API Key
3. Copy the key (starts with inv_sk_) — it's shown once

Setup

Claude Code

claude mcp add invompt -e INVOMPT_API_KEY=inv_sk_... -- npx invompt-mcp

Claude Desktop

Edit ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):

{
  "mcpServers": {
    "invompt": {
      "command": "npx",
      "args": ["-y", "invompt-mcp"],
      "env": {
        "INVOMPT_API_KEY": "inv_sk_..."
      }
    }
  }
}

Restart Claude Desktop after saving.

Cursor

Add to ~/.cursor/mcp.json (global) or .cursor/mcp.json (project):

{
  "mcpServers": {
    "invompt": {
      "command": "npx",
      "args": ["-y", "invompt-mcp"],
      "env": {
        "INVOMPT_API_KEY": "inv_sk_..."
      }
    }
  }
}

VS Code (Copilot)

Add to .vscode/mcp.json:

{
  "servers": {
    "invompt": {
      "type": "stdio",
      "command": "npx",
      "args": ["-y", "invompt-mcp"],
      "env": {
        "INVOMPT_API_KEY": "inv_sk_..."
      }
    }
  }
}

VS Code uses "servers", not "mcpServers".

Windsurf

Add to ~/.codeium/windsurf/mcp_config.json:

{
  "mcpServers": {
    "invompt": {
      "command": "npx",
      "args": ["-y", "invompt-mcp"],
      "env": {
        "INVOMPT_API_KEY": "inv_sk_..."
      }
    }
  }
}

Cline

Open Cline sidebar → MCP Servers → Configure, then add:

{
  "mcpServers": {
    "invompt": {
      "command": "npx",
      "args": ["-y", "invompt-mcp"],
      "env": {
        "INVOMPT_API_KEY": "inv_sk_..."
      }
    }
  }
}

Amazon Q Developer

Add to ~/.aws/amazonq/mcp.json:

{
  "mcpServers": {
    "invompt": {
      "command": "npx",
      "args": ["-y", "invompt-mcp"],
      "env": {
        "INVOMPT_API_KEY": "inv_sk_..."
      }
    }
  }
}

Codex (OpenAI)

Add to ~/.codex/config.toml:

[mcp_servers.invompt]
command = "npx"
args = ["-y", "invompt-mcp"]

[mcp_servers.invompt.env]
INVOMPT_API_KEY = "inv_sk_..."

Bun users

Replace npx with bunx and remove -y in any configuration above:

"command": "bunx",
"args": ["invompt-mcp"],

Usage

Once connected, ask your AI assistant:

"Create an invoice for 40 hours of web development at $150/hr for Acme Corp, due in 30 days"

The assistant reads the format spec, generates the invoice data, calls the API, and returns a link to your invoice.

How it works

You describe the invoice in plain English
  → LLM reads invompt://spec/iml/v1 to learn the format
  → LLM generates invoice data as YAML
  → LLM calls create_invoice
  → Invompt validates, renders, and stores the invoice
  → You get a shareable URL

Templates

Pass templateId to choose a look:

Environment variables

The resource and prompt work without an API key — only the tool needs one.

Development

npm install       # install dependencies
npm run build     # compile TypeScript
npm test          # run tests
npm run dev       # run from source

License

MIT — see LICENSE