Skip to main content

Developing with AI and MCP

SECUTIX exposes an MCP (Model Context Protocol) server so AI coding assistants and agents — such as Claude Code — can browse the API documentation and generate code directly from a chat session, without leaving the tool you're already using.

1. Request an access token​

MCP calls are authenticated the same way as documentation access. If you don't already have a token:

  1. Go to /request-access and submit your details.
  2. Once approved, you'll receive an email with a magic link and your access token shown as a separate, copy-pasteable string.

The same token works for both the documentation site and the MCP server — you only need one.

2. Add the MCP server to your client​

For Claude Code, run:

claude mcp add --transport http secutix-mcp-server https://developer.api.quantum.secutix.com/mcp-server/mcp \
  --header "stx-api-access: YOUR_TOKEN"

Replace YOUR_TOKEN with the token from your approval email.

Every call to the MCP server must include the stx-api-access header — requests without it, or with an invalid or expired token, are rejected before they reach the server.

Keeping the token out of shell history​

Instead of pasting the token directly into the claude mcp add command, you can store it as an environment variable and reference it from your MCP configuration (.mcp.json or ~/.claude.json):

{
  "mcpServers": {
    "secutix-mcp-server": {
      "type": "http",
      "url": "https://developer.api.quantum.secutix.com/mcp-server/mcp",
      "headers": {
        "stx-api-access": "${STX_API_TOKEN}"
      }
    }
  }
}
export STX_API_TOKEN="YOUR_TOKEN"

3. Lost or need a new token?​

Resubmit the form at /request-access with the same email address — since you're already approved, a new token is generated and emailed immediately, no re-approval needed.