Unblu

Provides token-efficient access to Unblu API endpoints via an MCP server with progressive discovery and safety controls.

javascript

GitHub Stars

javascript

Language

5 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": {
    "detailobsessed-unblu-mcp": {
      "command": "unblu-mcp",
      "args": [
        "--provider",
        "k8s",
        "--environment",
        "dev"
      ],
      "env": {
        "PATH": "/Users/YOUR_USERNAME/.local/bin:/opt/homebrew/bin:/usr/local/bin:/usr/bin:/bin",
        "UNBLU_API_KEY": "your-api-key",
        "UNBLU_BASE_URL": "https://your-instance.unblu.cloud/app/rest/v4"
      }
    }
  }
}

You can run an MCP server that exposes Unblu REST API endpoints to your AI assistants with progressive discovery, safety controls, and Kubernetes-friendly port-forwarding. This makes it easy to access 300+ Unblu endpoints efficiently while keeping control over what your agents can do.

How to use

You use an MCP client (such as Claude Desktop or Windsurf) to connect to the Unblu MCP server. Start the server locally or in your Kubernetes cluster, then configure your client to point at it. The MCP server presents a compact discovery surface and allows you to call any Unblu REST API endpoint through safe, policy-enforced operations.

How to install

Prerequisites: ensure you have the MCP tooling ready to install and run MCP servers in your environment.

uv tool install unblu-mcp

MCP client configuration examples

Add the Unblu MCP server configuration to your MCP client config. These examples show how to run the server locally with a Kubernetes provider or with direct API access using environment variables.

{
  "mcpServers": {
    "unblu": {
      "command": "unblu-mcp",
      "args": ["--provider", "k8s", "--environment", "dev"],
      "env": {
        "PATH": "/Users/YOUR_USERNAME/.local/bin:/opt/homebrew/bin:/usr/local/bin:/usr/bin:/bin"
      }
    }
  }
}

{
  "mcpServers": {
    "unblu": {
      "command": "unblu-mcp",
      "env": {
        "PATH": "/Users/YOUR_USERNAME/.local/bin:/opt/homebrew/bin:/usr/local/bin:/usr/bin:/bin",
        "UNBLU_BASE_URL": "https://your-instance.unblu.cloud/app/rest/v4",
        "UNBLU_API_KEY": "your-api-key"
      }
    }
  }
}

Configuration and helpers

The server supports a Kubernetes-based provider and direct API access through environment variables. Use the provided command and environment variable patterns to configure your client.

When you run with Kubernetes, the server will handle port-forwarding and auto-restart as needed.

Safety and behavior notes

Leverage the built-in safety controls and optional policy enforcement to regulate what your AI assistants can do against Unblu’s API surface.

Troubleshooting

If you encounter connectivity or policy issues, verify the MCP server is running, the client is configured with the correct command or URL, and that any required API keys or base URLs are correctly set in your environment.

Notes

The MCP server exposes a wide range of Unblu REST API v4 endpoints through a progressive discovery surface, enabling token-efficient access for AI assistants.

Available tools

list_services

List all API service categories available in the MCP server.

list_operations

List operations within a specific service category.

search_operations

Search operations by keyword across the API surface.

get_operation_schema

Retrieve the full schema for a specific operation.

call_api

Execute a selected API operation with provided parameters.

Built by

VeilStrat

AI signals for GTM teams