Docker MCP Registry

The Vectra AI MCP Server exposes several tool categories that allow AI assistants (like Claude, Cursor, or Windsurf) to interact with your security environment:

• Detection Management: List, filter, and get details for security detections; retrieve detection PCAP files; count detections; and mark detections as fixed or unfixed.
• Entity Management: List and query host and account entities; get detailed information by ID, name, or IP address; filter by state, priority, or tags; and access detection history.
• Investigation & Response: Create/manage investigation assignments; add investigation notes to entities; list assignments by user or entity; and manage lockdown entities.
• Platform Administration: List platform users (with role filtering) and manage user assignments/permissions; access platform configuration and logs.
• Advanced Analytics: Generate AI-driven detection summaries and visualize entity detection relationships via interactive graphs.

Using the Okta MCP Server involves specific security and privacy implications that users must understand:

Data Flow & Privacy

When a request is made, the interaction happens directly between the LLM and the Okta MCP tools. All data returned (e.g., complete user profiles, group memberships) is sent to and stored in the LLM's context for the duration of the conversation. Ensure you are comfortable with your Okta data being processed by your chosen LLM provider (e.g., Anthropic, OpenAI, Google).

HTTP Transport Security Warning

HTTP transport modes (Streamable HTTP and SSE) are high-risk. They can open unauthenticated HTTP servers with full access to your Okta tenant. Anyone reaching the network port can issue commands.

• Best Practice: Use the STDIO transport method (default) unless you have specific security controls in place.
• Warning: Using remote HTTP access via mcp-remote is considered EXTREMELY DANGEROUS.

Docker images are available for all transport types. Use the appropriate image based on your deployment needs.

STDIO (Recommended for Claude Desktop):

{
  "mcpServers": {
    "okta-mcp-server": {
      "command": "docker",
      "args": [
        "run", "-i", "--rm",
        "-e", "OKTA_CLIENT_ORGURL",
        "-e", "OKTA_API_TOKEN",
        "fctrid/okta-mcp-server:stdio"
      ],
      "env": {
        "OKTA_CLIENT_ORGURL": "https://your-org.okta.com",
        "OKTA_API_TOKEN": "your_api_token"
      }
    }
  }
}

Streamable HTTP (Current Standard):

docker run -d --name okta-mcp-http \
  -p 3000:3000 \
  -e OKTA_API_TOKEN=your_api_token \
  -e OKTA_CLIENT_ORGURL=https://your-org.okta.com \
  fctrid/okta-mcp-server:http

Building Images Locally:

docker build --target stdio -t okta-mcp-server:stdio .
docker build --target http -t okta-mcp-server:http .
docker build --target sse -t okta-mcp-server:sse .

To run the Wayfound MCP server example locally, follow these steps to clone the repository, install dependencies, and configure your environment variables.

Prerequisites

• Python 3.10 or higher
• A Wayfound MCP API key
• An OpenAI API key

Installation Steps

1. Clone the repository:

   git clone https://github.com/Wayfound-AI/wayfound-mcp-example
   cd wayfound-mcp-example

2. Install dependencies:

   pip install -r requirements.txt

3. Configure environment variables: Create a .env file in the project root and add your keys:

   WAYFOUND_MCP_API_KEY=your_mcp_api_key_here
   OPENAI_API_KEY=your_openai_api_key_here

4. Run the example:

   python main.py

git clone https://github.com/Wayfound-AI/wayfound-mcp-example
cd wayfound-mcp-example
pip install -r requirements.txt
# Create .env file with WAYFOUND_MCP_API_KEY and OPENAI_API_KEY
python main.py