Enkrypt AI

Integrates real-time prompt risk analysis, red-teaming, and AI safety tooling into MCP clients.

python

3

GitHub Stars

python

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": {
    "enkryptai-enkryptai-mcp-server": {
      "command": "uv",
      "args": [
        "--directory",
        "PATH/TO/enkryptai-mcp-server",
        "run",
        "src/mcp_server.py"
      ],
      "env": {
        "ENKRYPTAI_API_KEY": "YOUR ENKRYPTAI API KEY"
      }
    }
  }
}

You can run the Enkrypt AI MCP Server to analyze prompts, perform red-teaming, and monitor AI safety directly within any MCP-compatible client. This server lets you inspect prompts in real time, simulate adversarial inputs, and integrate safety tooling into your assistant-driven workflows.

How to use

Connect the Enkrypt AI MCP Server to your MCP client to enable on-device prompt analysis and safety tooling. Once configured, your MCP client will be able to invoke the server to perform real-time risk analysis, generate adversarial prompts for red-teaming, and monitor tool-based LLM activity within your workflow. Ensure you provide your API key so the server can access the Enkrypt services.

How to install

Prerequisites you need before installing: a working Python environment and the uv tool installed on your machine.

# Prerequisite: ensure uv is installed
# Then install the MCP server package in editable mode
uv pip install -e .

Configuration and usage notes

The server is designed to connect to MCP clients like Cursor or Claude Desktop. You will run the MCP server locally and point your client at it using the provided configuration blocks.

{
  "mcpServers": {
    "EnkryptAI-MCP": {
      "command": "uv",
      "args": [
        "--directory",
        "PATH/TO/enkryptai-mcp-server",
        "run",
        "src/mcp_server.py"
      ],
      "env": {
        "ENKRYPTAI_API_KEY": "YOUR ENKRYPTAI API KEY"
      }
    }
  }
}

Examples for MCP clients

Cursor and Claude Desktop can both use the same MCP server config pattern to start the Enkrypt AI MCP Server. The examples below show what to paste into each client’s MCP settings.

{
  "mcpServers": {
    "EnkryptAI-MCP": {
      "command": "uv",
      "args": [
        "--directory",
        "PATH/TO/enkryptai-mcp-server",
        "run",
        "src/mcp_server.py"
      ],
      "env": {
        "ENKRYPTAI_API_KEY": "YOUR ENKRYPTAI API KEY"
      }
    }
  }
}

Available tools

prompt_risk_analysis

Performs real-time risk analysis on prompts to identify potential safety issues or jailbreak attempts.

red_team_adversarial_generation

Generates adversarial prompts to test and strengthen safety boundaries via red-teaming scenarios.

llm_monitoring

Monitors tool-based LLM behavior within MCP clients to detect unsafe usage patterns and ensure compliance with safety tooling.

mcp_integration

Provides seamless integration with MCP clients like Cursor and Claude Desktop for unified safety workflows.