Portkey

A Model Context Protocol (MCP) server implementation for the Portkey AI platform, providing a standardized interface for managing AI configurations, workspaces, and analytics.
  • typescript

3

GitHub Stars

typescript

Language

6 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs
{
  "mcpServers": {
    "r-huijts-portkey-admin-mcp-server": {
      "command": "node",
      "args": [
        "/path/to/portkey-server/build/index.js"
      ],
      "env": {
        "PORTKEY_API_KEY": "your_portkey_api_key_here"
      }
    }
  }
}

You deploy Portkey MCP Server to connect Claude with Portkey’s platform, enabling centralized management of configurations, workspaces, analytics, and user access through an MCP client.

How to use

You use the Portkey MCP Server to access Portkey’s management capabilities from your MCP client. Start the local server, then configure your MCP client to point at the local stdio endpoint. This lets you perform user administration, workspace management, analytics reporting, and API configuration tasks from a single, unified interface.

Typical usage patterns include listing users and inviting new ones, creating and configuring workspaces, viewing usage analytics and costs, monitoring API keys and rate limits, and managing access control across your organization. You can also run maintenance tasks for configurations, review routing and retry policies, and generate reports across time ranges.

How to install

Prerequisites: you need Node.js and npm installed on your system.

From Source steps describe how to set up the server locally and prepare it for use with Claude Desktop.

Step by step commands you will run are listed below.

# 1) Clone the repository
# (Run this in your preferred directory)

# 2) Install dependencies
npm install

# 3) Copy the example environment file
cp .env.example .env

# 4) Add your Portkey API key to the .env file
# Open .env and set PORTKEY_API_KEY=your_portkey_api_key_here

# 5) Update Claude configuration to include the Portkey MCP Server
# The following shows where to place the runtime config
{
  "mcpServers": {
    "portkey-server": {
      "command": "node",
      "args": [
        "/path/to/portkey-server/build/index.js"
      ],
      "env": {
        "PORTKEY_API_KEY": "your_portkey_api_key_here"
      }
    }
  }
}

Additional installation paths

If you prefer automated installation, you can install Portkey MCP Server for Claude Desktop automatically using Smithery.

npx -y @smithery/cli install @r-huijts/portkey-admin-mcp-server --client claude

Notes on configuration and environment

Environment variable required by the server: PORTKEY_API_KEY. This key is used to authenticate requests to Portkey’s API and should be kept secure.

Examples of real-world use cases

  • View current API usage statistics across models
  • Generate analytics reports for recent periods
  • Create a new workspace for a team project
  • Review current API key usage and remaining credits
  • Set up rate limiting and fallback behavior for production endpoints
  • Manage team access and workspace memberships
  • Inspect latency statistics for API calls
  • Configure custom headers for API requests
  • List and review workspace configurations and virtual keys
  • Track request metrics, token usage, and costs over time

Available tools

User Administration

List and manage all users in your Portkey organization, including invitations and role assignments.

Workspace Management

Create, configure, and monitor workspaces; manage user access and workspace settings.

Analytics & Reporting

Access usage analytics, performance metrics, cost analysis, and filtered reports.

Virtual Key Management

Manage API keys with usage limits and rate limits; monitor status and configurations.

Configuration & API Settings

Browse configurations, cache settings, retry policies, routing strategies, and API endpoints.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational