---
name: orchardcore-ai-mcp-client
description: Skill for configuring CrestApps MCP client connections in Orchard Core. Covers SSE and local client transports, authentication, recipes, and assigning MCP connections to AI experiences. Use this skill when requests mention Orchard Core MCP Client, Configure MCP Clients, Feature Overview, Enable the SSE MCP Client, Add a Remote MCP Connection via Admin UI, SSE Authentication Types, or closely related Orchard Core implementation, setup, extension, or troubleshooting work. Strong matches include work with CrestApps.OrchardCore.AI.Mcp, CrestApps.OrchardCore.AI.Mcp.LocalClient, CrestApps.OrchardCore.AI, ApiKey, OAuth2ClientCredentials, OAuth2PrivateKeyJwt, OAuth2MutualTls, CustomHeaders. It also helps with Add a Remote MCP Connection via Admin UI, SSE Authentication Types, SSE Recipe Example, plus the code patterns, admin flows, recipe steps, and referenced examples captured in this skill.
license: Apache-2.0
metadata:
  author: CrestApps Team
  version: "1.0"
---

# Orchard Core MCP Client - Prompt Templates

## Configure MCP Clients

You are an Orchard Core expert. Generate admin, configuration, and recipe guidance for connecting Orchard Core to external MCP servers through the CrestApps MCP client features.

### Guidelines
- Use `CrestApps.OrchardCore.AI.Mcp` for remote MCP servers over SSE.
- Use `CrestApps.OrchardCore.AI.Mcp.LocalClient` for local MCP servers over standard input/output.
- Remote SSE connections are managed under **Artificial Intelligence → MCP Connections**.
- Sensitive secrets are encrypted at rest using ASP.NET Core Data Protection.
- Do not export encrypted values and do not commit plaintext secrets.

### Feature Overview

| Transport | Feature ID | Description |
|-----------|-----------|-------------|
| SSE | `CrestApps.OrchardCore.AI.Mcp` | Connect to remote MCP servers over HTTP/SSE |
| Local Client | `CrestApps.OrchardCore.AI.Mcp.LocalClient` | Connect to local MCP servers over stdin/stdout |

### Enable the SSE MCP Client

```json
{
  "steps": [
    {
      "name": "Feature",
      "enable": [
        "CrestApps.OrchardCore.AI",
        "CrestApps.OrchardCore.AI.Mcp"
      ],
      "disable": []
    }
  ]
}
```

### Add a Remote MCP Connection via Admin UI

1. Navigate to **Artificial Intelligence → MCP Connections**.
2. Click **Add Connection**.
3. Under **Server Sent Events (SSE)**, click **Add**.
4. Enter:
   - **Display Text**
   - **Endpoint** such as `https://localhost:1234/`
   - **Authentication** method
5. Save the connection.

### SSE Authentication Types

| Type | Description |
|------|-------------|
| `Anonymous` | No authentication |
| `ApiKey` | Sends an API key using a configurable header and optional prefix |
| `Basic` | Uses HTTP basic authentication |
| `OAuth2ClientCredentials` | Uses OAuth 2.0 client credentials |
| `OAuth2PrivateKeyJwt` | Uses a private key JWT client assertion |
| `OAuth2MutualTls` | Uses mTLS with a client certificate |
| `CustomHeaders` | Sends raw headers as JSON |

### SSE Recipe Example

```json
{
  "steps": [
    {
      "name": "McpConnection",
      "connections": [
        {
          "DisplayText": "Example server",
          "Properties": {
            "SseMcpConnectionMetadata": {
              "Endpoint": "https://localhost:1234/",
              "AuthenticationType": "ApiKey",
              "ApiKeyHeaderName": "Authorization",
              "ApiKeyPrefix": "Bearer",
              "ApiKey": "your-api-key-here"
            }
          }
        }
      ]
    }
  ]
}
```

### Enable the Local MCP Client

```json
{
  "steps": [
    {
      "name": "Feature",
      "enable": [
        "CrestApps.OrchardCore.AI",
        "CrestApps.OrchardCore.AI.Mcp",
        "CrestApps.OrchardCore.AI.Mcp.LocalClient"
      ],
      "disable": []
    }
  ]
}
```

### Add a Local MCP Connection via Admin UI

1. Navigate to **Artificial Intelligence → MCP Connections**.
2. Click **Add Connection**.
3. Under **Standard Input/Output (Stdio)**, click **Add**.
4. Enter:
   - **Display Text**: `Global Time Capabilities`
   - **Command**: `docker`
   - **Command Arguments**: `["run", "-i", "--rm", "mcp/time"]`
5. Save the connection.

### Local Client Recipe Example

```json
{
  "steps": [
    {
      "name": "McpConnection",
      "connections": [
        {
          "DisplayText": "Global Time Capabilities",
          "Properties": {
            "StdioMcpConnectionMetadata": {
              "Command": "docker",
              "Arguments": [
                "run",
                "-i",
                "--rm",
                "mcp/time"
              ]
            }
          }
        }
      ]
    }
  ]
}
```

### Assign MCP Connections to AI Experiences

After creating a connection, assign it where needed:

- **AI Profiles**: select connections under the **Capabilities** tab.
- **AI Templates**: select connections under the **Capabilities** tab for profile-source templates.
- **Chat Interactions**: select connections under the **Capabilities** area of the interaction editor.

### Security Notes

- Sensitive values such as API keys, client secrets, private keys, and certificates are encrypted at rest.
- Sensitive values are not exported in deployment packages.
- Do not commit raw secrets in recipes.
- Use secure secret storage for production deployments.

### Best Practices

- Use SSE for remote MCP services shared across environments.
- Use Stdio for local tools, containers, and offline utilities.
- Give connections descriptive names so editors can choose correctly.
- Prefer standard authentication types over custom headers unless required.