System Prompt

A System Prompt is a high-level instruction set that establishes the "ground rules" for an AI model's behavior during a session. In MCP, it is a crucial component of the sampling context.

Role in MCP

When an MCP host initiates a sampling request, it often includes a systemPrompt that incorporates:

  1. Core Identity: "You are a helpful coding assistant."
  2. MCP Context: Information about the available tools and resources provided by the servers.
  3. Safety Constraints: Guidelines on data handling and user privacy.

Example in Sampling

{
  "method": "sampling/createMessage",
  "params": {
    "systemPrompt": "You have access to a GitHub server. Use the 'list_issues' tool to help the user.",
    "messages": [...]
  }
}

A well-crafted system prompt ensures the AI understands how to effectively use the specialized capabilities offered by the connected MCP servers.

Dynamic Prompt Enrichment with HasMCP

HasMCP simplifies the management of System Prompts by ensuring that tool-related context is always accurate and optimized. Through Automated OpenAPI Mapping, HasMCP provides the AI model with rich and semantically precise tool descriptions that are automatically injected into the system instructions. This reduces the burden on developers to manually maintain prompt templates and ensures that as backend APIs evolve, the AI’s understanding of its available capabilities remains perfectly in sync, leading to more reliable and efficient tool selection during every turn.

Questions & Answers

What is the primary role of a "System Prompt" in an AI session?

A system prompt establishes the "ground rules" for an AI model's behavior. it defines its identity (e.g., "You are a code reviewer"), its constraints, and provide foundational instructions that influence every subsequent interaction.

How does an MCP host use system prompts during sampling?

During a sampling request, the host incorporates task-specific context—including information about available server tools and resources—directly into the system prompt to guide the AI's reasoning and ensure it uses server capabilities correctly.

How does HasMCP automate the enrichment of system instructions?

HasMCP uses Automated OpenAPI Mapping to dynamically generate and inject precise tool descriptions into the system prompt. This ensures the AI model always has a current and accurate understanding of available capabilities without requiring manual developer updates.

Back to Glossary