Prompt a model connected to the Vault Radar MCP server

Use prompts with your AI model to gain insights into your Vault Radar data.

Overview

To trigger the MCP server tools, ask your LLM questions about your Vault Radar data sources, resources, or events. Note that for your LLM to use the server, your questions must pertain to data from your HCP Vault Radar instance. The LLM uses other sources, such as content from the internet, to respond to general questions about Vault Radar or requests to generate code.

Prerequisites

• Vault Radar MCP server deployed and configured with your client

Prompt your MCP client

Your model may still decide to use other resources to formulate responses, but you can use the following guidelines to help you create prompts that are more likely to be routed through the MCP server:

• Use specific keywords related to the tools available in the MCP server, such as "data sources", "resources", or "events".
• Ask questions that require the model to retrieve information from your HCP Vault Radar instance, such as "What data sources are available in my Vault Radar project?" or "What events have been detected in my Vault Radar project?".
• Avoid general questions that do not relate to your HCP Vault Radar instance, as the model may not use the MCP server tools for those queries.

Examples

You can use the following set of prompts as a guide you. Your responses may vary.

• "What are the top 10 most vulnerable data sources in my Vault Radar project?"
• "What are the top 20 riskiest resources in my Vault Radar project?"
• "What are the top 10 risks I should mitigate first?"
• "Which resources have the most critical events in my Vault Radar project?"
• "How many GitHub data sources do I have?"
• "Which of the leaked secrets are both critical severity and present in Vault?"
• "How many passwords are detected in my environment?"