SamplingDefinition

turul_mcp_protocol::sampling

Trait SamplingDefinition 

Source 
pub trait SamplingDefinition:
    HasSamplingConfig
    + HasSamplingContext
    + HasModelPreferences {
    // Provided method
    fn to_create_params(&self) -> CreateMessageParams { ... }
}
Expand description

Complete MCP Sampling Creation - Build AI model interaction and completion systems.

This trait represents a complete, working MCP sampling configuration that controls how AI models generate completions with precise parameter control, context management, and model preferences. When you implement the required metadata traits, you automatically get SamplingDefinition for free via blanket implementation.

§What You’re Building

A sampling configuration is an AI interaction system that:

  • Controls model generation parameters (temperature, tokens, etc.)
  • Manages conversation context and system prompts
  • Specifies model preferences and capabilities
  • Handles structured completion requests

§How to Create a Sampling Configuration

Implement these three traits on your struct:


// This struct will automatically implement SamplingDefinition!
struct CodeReviewSampling {
    review_type: String,
    language: String,
}

impl HasSamplingConfig for CodeReviewSampling {
    fn max_tokens(&self) -> u32 {
        2000 // Enough for detailed code reviews
    }

    fn temperature(&self) -> Option<f64> {
        Some(0.3) // Lower temperature for consistent code analysis
    }

    fn stop_sequences(&self) -> Option<&Vec<String>> {
        None // No special stop sequences needed
    }
}

impl HasSamplingContext for CodeReviewSampling {
    fn messages(&self) -> &[SamplingMessage] {
        // Static messages for this example
        &[]
    }

    fn system_prompt(&self) -> Option<&str> {
        Some("You are an expert code reviewer. Analyze the provided code for bugs, performance issues, and best practices.")
    }

    fn include_context(&self) -> Option<&str> {
        Some(&self.language) // Include programming language context
    }
}

impl HasModelPreferences for CodeReviewSampling {
    fn model_preferences(&self) -> Option<&ModelPreferences> {
        None // Use default model
    }
}

// Now you can use it with the server:
let sampling = CodeReviewSampling {
    review_type: "security".to_string(),
    language: "rust".to_string(),
};

// The sampling automatically implements SamplingDefinition
let create_params = sampling.to_create_params();

§Key Benefits

  • Precise Control: Fine-tune model behavior for specific tasks
  • Context Management: Rich conversation and system prompt support
  • Model Flexibility: Support for different AI models and capabilities
  • MCP Compliant: Fully compatible with MCP 2025-06-18 specification

§Common Use Cases

  • Code review and analysis systems
  • Creative writing assistance
  • Technical documentation generation
  • Question-answering with domain context
  • Conversational AI with specific personalities

Provided Methods§

Source

fn to_create_params(&self) -> CreateMessageParams

Convert to CreateMessageParams

Implementors§