Creating and Managing Your MCP Copilot

This document provides a step-by-step approach to adding an MCP server, creating an MCP copilot project, and configuring the necessary settings to personalize your application. Follow these instructions to effectively manage your project resources and customize your AI assistant.

Step 1

To begin, click the menu icon to access navigation options. Proceed to the 'Manage MCPs' section in 'Tools Management' to add an MCP.

Click the 'Add to MCP Library' button to include the required MCP server.

Step 2

Enter the credentials for the MCP server. Select the groups with whom you wish to share access, and then click 'Add'. Your MCP will now be visible in 'My MCP Library'.

Step 3

After adding the MCP server, navigate to 'Projects' to create an MCP copilot. Click 'Create Project' and enter the project name to initiate your MCP agent creation.

Step 4

Select 'Private' to restrict project access to yourself, or choose 'Shared' to allow team collaboration. Specify the groups that should have access to this project.

Step 5

Select 'MCP Copilot' to develop your AI-powered copilot agent. Click 'Create and Continue' to activate your new MCP copilot project.

Step 6

Choose the MCP data source to connect with your Copilot. In this demonstration, we'll use GitHub as the data source. Click 'Next' to configure your copilot's GitHub source.

Step 7

Select your AI model according to your requirements, then click the 'Add' button to begin the setup. Choose your preferred model version from the available options.Click 'Next' to proceed with model configuration.

Step 8

Expand 'Advanced Settings' for additional customization options.

Step 9

Enhance your AI's user journey capabilities by adding a custom chat reasoning.

Step 10

Define your chat reasoning text and click 'Next' to continue. Use prompt personalization to optimize application efficiency by adjusting prompts and parameters.

Step 11

Customize the system prompt to guide your AI's personality and behavior.

Step 12

Modify content filter prompts to ensure interactions are safe and tailored to your standards. Toggle filters to manage moderation for sensitive content, adjusting filtering of restricted terms as necessary.

Step 13

Add a prompt to generate smart, contextual conversation names automatically. Customize follow-up question prompts to foster deeper engagement. Enable follow-up questions for an interactive AI experience.

Step 14

Choose an AI model for follow-up question generation. Include an organization policy prompt to align responses with company rules. Enable policy checks for compliant AI interactions.

Step 15

Select the optimal AI model for effective policy enforcement. Click 'Next' to proceed with your custom prompt configurations and adjust your application card details.

Step 16

Enter a name for your application to personalize its app card. Choose a category to facilitate app discovery, and set its accessibility to 'Public' for wider availability.

Step 17

Select 'Shared' to limit app access to specific groups, or choose 'Private' for exclusive access. Click 'Next' to save changes and proceed with the app setup.

Step 18

Customize your application's branding. Click here to update the browser tab icon and make your app stand out. Upload a unique brand logo to represent your application.

Step 19

Replace the central icon to better align your app with your brand's aesthetic. Select the desired resource size to optimize project performance. Click 'Deploy' to publish your branding and updates across the application.

Step 20

Click 'Next' to continue with AI development customization. Test your chat assistants and fine-tune parameters before deployment.

Step 21

Refine various parameters for optimal performance. Test your chatbot to verify settings before deployment. Here's an example of our chatbot.

Step 22

Explore options on your application card. Access API settings for your AI project, and open the 'Create API Token' tab to generate secure access credentials.

Step 23

Assign a unique name to your new API token for organizational clarity. Click 'Create Token' to generate a new API token for your project. Switch to the 'Existing API Token' tab to manage existing tokens.

Step 24

Consult the API usage documentation for detailed integration guidance. Click 'View Documentation' for examples and best practices. Select 'Authorize' to authenticate and access API endpoints securely.

Step 25

Enter your API token in the 'Value' field to authenticate. Expand the '/api/v1/predict' endpoint for prediction details. Click 'Try it out' to test this endpoint with your data.

Step 26

Click 'Execute' to view real-time API responses based on your input. Review the project's activity logs for performance monitoring and issue detection. Click 'Edit' to modify chatbot settings.

Step 27

Test your chatbot by selecting 'Try this chatbot'.