Connector Setup

Safeguard hosts its MCP server, so there is nothing to install:

• MCP endpoint (Streamable HTTP): https://mcp.safeguard.sh/mcp
• Claude connector URL: https://mcp.safeguard.sh/mcp/anthropic
• Sign-in: OAuth 2.0. Your client opens Safeguard's sign-in page: sign in, check the app name and the host the page says the authorization code will be sent to, then click Authorize access.
• MCP key, only for a client that can only send a fixed header: in Safeguard, open Settings → Developer, create a key of the type MCP key, and send it as Authorization: Bearer sg_sk_.... If you don't see the Developer tab or the MCP key type, ask your Safeguard administrator: API-key access to MCP is turned on per workspace.

There is no SSE endpoint, and the server never reads environment variables such as SAFEGUARD_API_KEY from your machine. Safeguard does not publish an npm package or a local bridge for its MCP server; treat any package that claims to be one as untrusted.

Claude (claude.ai and Claude Desktop)

1. In Claude, open Customize → Connectors, click + Add, then Add custom connector.
2. Enter the name Safeguard Security and the URL https://mcp.safeguard.sh/mcp/anthropic, then click Continue.
3. Claude detects that Safeguard signs in with OAuth; click Continue. Under Authentication, choose Sign in now or Sign in when needed. Under OAuth client, choose Register automatically: Safeguard accepts only client IDs it issued itself. Click Add.
4. Sign in to Safeguard, check the app name and host, then click Authorize access.

On a Team or Enterprise plan, an Owner or Primary Owner adds the connector under Organization settings → Connectors first; members then connect it from Customize → Connectors. Claude Desktop uses the same connectors, so there is nothing to add to claude_desktop_config.json. Full guide: Claude.

Claude Code

claude mcp add --transport http safeguard https://mcp.safeguard.sh/mcp

Then run /mcp in Claude Code and follow the steps in your browser to sign in. To use an MCP key instead, add --header "Authorization: Bearer sg_sk_..." to the command.

ChatGPT

1. In ChatGPT, open Plugins, select the plus button, then Add custom MCP server.
2. Enter a name and description, and under Connection enter https://mcp.safeguard.sh/mcp.
3. Configure authentication with OAuth (ChatGPT registers itself; there is no client ID to copy), review the risk warning, select I understand and want to continue, then Create as a plugin.
4. Sign in to Safeguard, check the app name and host, then click Authorize access.

Account and workspace policies apply to custom MCP servers in ChatGPT. A Custom GPT Action that imports https://mcp.safeguard.sh/openapi.json with an MCP key is an alternative: see ChatGPT.

For the OpenAI API, add an mcp tool to a Responses API request with server_url set to https://mcp.safeguard.sh/mcp. Safeguard needs a credential on every request; OpenAI's MCP and connectors guide describes how to pass one.

Cursor

Add this to ~/.cursor/mcp.json (all projects) or .cursor/mcp.json (one project):

{
  "mcpServers": {
    "safeguard": {
      "url": "https://mcp.safeguard.sh/mcp"
    }
  }
}

Cursor opens Safeguard's sign-in page when it connects. Full guide: Cursor.

VS Code (GitHub Copilot)

Add this to .vscode/mcp.json, or run MCP: Open User Configuration to add it for every workspace:

{
  "servers": {
    "safeguard": {
      "type": "http",
      "url": "https://mcp.safeguard.sh/mcp"
    }
  }
}

On the first connection VS Code opens a browser window to sign in.

Gemini CLI

gemini mcp add --transport http safeguard https://mcp.safeguard.sh/mcp/gemini

Or add "safeguard": { "httpUrl": "https://mcp.safeguard.sh/mcp/gemini" } under mcpServers in ~/.gemini/settings.json. Gemini CLI discovers Safeguard's OAuth sign-in when it first connects. Full guide: Gemini.

Other clients

Add a remote MCP server (Streamable HTTP) with the URL https://mcp.safeguard.sh/mcp. If the client supports OAuth for MCP servers, it opens Safeguard's sign-in page; otherwise send an MCP key in the Authorization header. Per-assistant guides are listed under MCP Integrations.

Verifying the Connection

In any client, ask: "What Safeguard tools do you have?"

A working connection lists tools such as safeguard_list_projects and safeguard_list_vulnerabilities. A new workspace starts with six read-only tools, and an administrator turns on more (see Tool availability & gating). If the list is empty, see FAQ & Troubleshooting on the MCP index page.

Related

• MCP Server overview — the full tool surface and feature matrix.
• API Reference — direct REST API if you prefer not to use MCP.
• Desktop Application — pairs well with any MCP-aware client.