leximo-ai-call-assistant-mcp-server
Official# Leximo AI Call Assistant — MCP Server
[](https://www.npmjs.com/package/leximo-ai-call-assistant-mcp-server)
[](https://nodejs.org)
[](https://modelcontextprotocol.io)
[](LICENSE)
An [MCP (Model Context Protocol)](https://modelcontextprotocol.io) server that lets you schedule AI phone calls and manage [Leximo](https://leximo.ai) assignments directly from Claude Desktop or Claude Code — no app switching needed.
---
## Quick Install
### Claude Code (one command)
```bash
claude mcp add leximo -e LEXIMO_API_TOKEN=your-token -- npx -y leximo-ai-call-assistant-mcp-server
```
Replace `your-token` with your API token from [concierge.leximo.ai/profile](https://concierge.leximo.ai/profile).
### Claude Code (plugin marketplace)
```
/plugin marketplace add leximo-ai/leximo-ai-call-assistant-mcp-server
```
Then install the plugin:
```
/plugin install leximo-ai-call-assistant
```
---
## Manual Setup
### 1. Get your API token
1. Go to [concierge.leximo.ai](https://concierge.leximo.ai) and sign in
2. Open your [profile page](https://concierge.leximo.ai/profile)
3. Copy your JWT access token
### 2. Configure Claude Desktop
Add to `~/Library/Application Support/Claude/claude_desktop_config.json` (macOS) or `%APPDATA%\Claude\claude_desktop_config.json` (Windows):
```json
{
"mcpServers": {
"leximo": {
"command": "npx",
"args": ["-y", "leximo-ai-call-assistant-mcp-server"],
"env": {
"LEXIMO_API_TOKEN": "your-token"
}
}
}
}
```
### 3. Configure Claude Code (manual)
Add to your Claude Code MCP settings:
```json
{
"mcpServers": {
"leximo": {
"command": "npx",
"args": ["-y", "leximo-ai-call-assistant-mcp-server"],
"env": {
"LEXIMO_API_TOKEN": "your-token"
}
}
}
}
```
---
## Features
- **Assignments** — Create, list, view, and delete AI phone call assignments
- **AI Agents** — Browse available calling agents and pick the right one
- **Task Proposals** — Get AI-generated improvements for your call instructions
- **Credits** — Check your credit balance and usage history
- **Subscriptions** — View your plan, browse available plans, and subscribe
- **Notifications** — View call completion events and system notifications
---
## Available Tools
| Tool | Description |
|------|-------------|
| `get_profile` | Get your user profile and account details |
| `get_credits` | Check credit balance, usage history, and subscription summary |
| `get_subscription` | View active subscription details |
| `get_plans` | List available subscription plans with pricing |
| `create_checkout_session` | Get a checkout URL to subscribe to a plan |
| `list_agents` | List available AI calling agents |
| `get_agent` | Get details of a specific agent |
| `list_assignments` | List all your assignments (paginated) |
| `get_assignment` | View a specific assignment with results and transcript |
| `create_assignment` | Create a new phone call assignment |
| `delete_assignment` | Delete an assignment |
| `get_assignment_proposals` | Get AI suggestions to improve your task description |
| `list_notifications` | Get call completion events and notifications |
---
## Example Prompts
Once configured, ask Claude things like:
- _"Show me my Leximo assignments"_
- _"Create a call to +1234567890 to book a restaurant for 2 at 7pm Friday"_
- _"How many credits do I have left?"_
- _"What subscription plans are available?"_
- _"Show me the transcript from my last call"_
- _"What agents are available and which one is best for restaurant bookings?"_
---
## Development
```bash
npm install
npm run build # Compile TypeScript
npm start # Run compiled server
npm run dev # Run with tsx (hot reload)
```
### Test with MCP Inspector
```bash
LEXIMO_API_TOKEN=your-token npx @modelcontextprotocol/inspector node dist/index.js
```
### Environment Variables
| Variable | Required | Description |
|----------|----------|-------------|
| `LEXIMO_API_TOKEN` | Yes | JWT token from [concierge.leximo.ai/profile](https://concierge.leximo.ai/profile) |
Copy `.env.example` to `.env` for local development.
---
## License
[MIT](LICENSE) © [Leximo](https://leximo.ai)
TDQS
Scored across 13 tools
All 13 tools have clearly distinct purposes. There is no overlap; each tool targets a unique resource or action, such as creating vs. listing assignments, retrieving agents, or managing subscriptions.
All tool names follow a consistent verb_noun pattern in snake_case (e.g., create_assignment, list_agents, get_credits). No mixing of conventions or ambiguous verbs.
With 13 tools, the set is well-scoped for the server's purpose (AI calling assignments and subscription management). Each tool has a clear role, and there are no extraneous tools.
The tool surface covers core workflows: assignment CRUD (except update), agent management, subscription management, and notifications. Missing an update_assignment tool is a minor gap, but the design intentionally uses creation via proposals.