Client setup

Once riskmanaged-mcp is installed, point your AI client at it. Every client below uses the same stdio command and the same two environment variables — only the config file location and its shape differ. Find your client and follow its section.

Claude Desktop#

Configuration file#

Configuration#

{
  "mcpServers": {
    "riskmanaged": {
      "command": "riskmanaged-mcp",
      "env": {
        "RISKMANAGED_TOKEN": "your_api_token_here",
        "RISKMANAGED_URL": "https://agent.riskmanaged.io"
      }
    }
  }
}

Steps#

1. Install riskmanaged-mcp: pip install riskmanaged-mcp
2. Open Claude Desktop → Settings → Developer → Edit Config
3. Paste the JSON above, replacing the token with your API token
4. Restart Claude Desktop
5. Look for the 🔨 hammer icon — click it to verify the RiskManaged tools appear

Cursor#

Configuration file#

~/.cursor/mcp.json

Configuration#

{
  "mcpServers": {
    "riskmanaged": {
      "command": "riskmanaged-mcp",
      "env": {
        "RISKMANAGED_TOKEN": "your_api_token_here",
        "RISKMANAGED_URL": "https://agent.riskmanaged.io"
      }
    }
  }
}

Steps#

1. Install riskmanaged-mcp: pip install riskmanaged-mcp
2. Create or edit ~/.cursor/mcp.json
3. Paste the configuration above
4. Restart Cursor
5. In the Cursor AI chat, type "Use the riskmanaged tools to list my strategies" to verify

Windsurf#

Configuration file#

Configuration#

{
  "mcpServers": {
    "riskmanaged": {
      "command": "riskmanaged-mcp",
      "env": {
        "RISKMANAGED_TOKEN": "your_api_token_here",
        "RISKMANAGED_URL": "https://agent.riskmanaged.io"
      }
    }
  }
}

Steps#

1. Install riskmanaged-mcp: pip install riskmanaged-mcp
2. Open Windsurf → Settings (Cmd/Ctrl + ,) → Cascade → Plugins → Edit configuration
3. Paste the configuration above
4. Open Command Palette (Cmd/Ctrl + Shift + P) → "Reload Window"
5. Verify in Cascade chat by looking for the MCP icon

VS Code#

Configuration#

VS Code supports MCP through Copilot's agent mode. Create a .vscode/mcp.json in your project:

{
  "servers": {
    "riskmanaged": {
      "command": "riskmanaged-mcp",
      "env": {
        "RISKMANAGED_TOKEN": "your_api_token_here",
        "RISKMANAGED_URL": "https://agent.riskmanaged.io"
      }
    }
  }
}

Steps#

1. Install riskmanaged-mcp: pip install riskmanaged-mcp
2. Create .vscode/mcp.json in your project root
3. Paste the configuration above
4. Open Copilot Chat → Switch to Agent mode
5. The RiskManaged tools will be available automatically

Cline#

Configuration#

Cline stores MCP config in cline_mcp_settings.json.

Steps#

1. Install riskmanaged-mcp: pip install riskmanaged-mcp

2. Open VS Code → Cline panel → MCP Servers icon → Configure tab

3. Click "Configure MCP Servers" to open the config file

4. Add the following configuration:

   Copy

   {
     "mcpServers": {
       "riskmanaged": {
         "command": "riskmanaged-mcp",
         "env": {
           "RISKMANAGED_TOKEN": "your_api_token_here",
           "RISKMANAGED_URL": "https://agent.riskmanaged.io"
         },
         "disabled": false
       }
     }
   }

5. Save and verify the server appears as connected in the Cline MCP panel

Continue.dev#

Configuration#

Continue.dev uses its own config file. Add MCP servers to your Continue configuration:

// In ~/.continue/config.json
{
  "mcpServers": [
    {
      "name": "riskmanaged",
      "command": "riskmanaged-mcp",
      "env": {
        "RISKMANAGED_TOKEN": "your_api_token_here",
        "RISKMANAGED_URL": "https://agent.riskmanaged.io"
      }
    }
  ]
}

Steps#

1. Install riskmanaged-mcp: pip install riskmanaged-mcp
2. Open Continue.dev settings (gear icon → config.json)
3. Add the mcpServers array entry above
4. Reload your editor

Hermes#

Prerequisites#

• Hermes agent installed (npm install -g hermes-agent or via the Nous Research installer)
• Python ≥ 3.11
• A RiskManaged API token (Profile → API Tokens)

Step 1: Install the MCP package#

curl -sSL https://riskmanaged.io/install.sh | bash

Or install manually:

# uv (fastest)
uv tool install riskmanaged-mcp@git+https://github.com/riskmanaged/riskmanaged-mcp.git

# pipx
pipx install git+https://github.com/riskmanaged/riskmanaged-mcp.git

Step 2: Add to Hermes config#

Edit ~/.hermes/config.yaml and add the riskmanaged entry under mcp_servers:

mcp_servers:
  riskmanaged:
    command: riskmanaged-mcp
    env:
      RISKMANAGED_TOKEN: "your_api_token_here"
      RISKMANAGED_URL: "https://riskmanaged.io"

Step 3: Reload#

In your Hermes terminal session, run:

/reload-mcp

No restart needed — Hermes hot-reloads MCP server connections.

Verify#

Ask Hermes: "What tools do you have available?" — you should see 30+ RiskManaged tools listed.

OpenClaw#

Prerequisites#

• OpenClaw daemon running (npm install -g openclaw@latest)
• Python ≥ 3.11
• A RiskManaged API token (Profile → API Tokens)

Option A: Chat command (easiest)#

Tell OpenClaw directly in your chat interface (Telegram, Discord, etc.):

install this skill: https://github.com/riskmanaged/riskmanaged-mcp

OpenClaw will clone the repo, read the AGENT.md skill metadata, and register the MCP tools automatically.

Option B: Manual skill installation#

1. Clone the repo into your OpenClaw skills/ directory:

   Copy

   cd ~/.openclaw/skills
   git clone https://github.com/riskmanaged/riskmanaged-mcp.git

2. Install the Python package:

   Copy

   curl -sSL https://riskmanaged.io/install.sh | bash

3. Set your API token in OpenClaw Settings → Skills → Workspace Skills → Environment Variables:

   • RISKMANAGED_TOKEN = your token
   • RISKMANAGED_URL = https://riskmanaged.io

Option C: MCPorter config#

If using MCPorter for MCP connections, add to your config:

{
  "mcpServers": {
    "riskmanaged": {
      "command": "riskmanaged-mcp",
      "env": {
        "RISKMANAGED_TOKEN": "your_api_token_here",
        "RISKMANAGED_URL": "https://riskmanaged.io"
      }
    }
  }
}

Verify#

Ask OpenClaw: "List my trading strategies" — it should call the RiskManaged list_strategies tool and return results.