mcp_kit.prompts.interpolation

Interpolation prompt engine for safe string substitution.

InterpolationPrompt

@dataclass
class InterpolationPrompt()

A prompt with interpolation text and optional default values.

Arguments:

• text: The prompt string with {placeholder} syntax
• defaults: Optional default values for placeholders

__post_init__

def __post_init__() -> None

Initialize defaults as empty dict if None.

InterpolationPromptEngine

class InterpolationPromptEngine(PromptEngine)

Prompt engine that performs safe string interpolation using predefined prompts.

This engine uses a map of prompt names to InterpolationPrompt objects with {placeholder} syntax for argument substitution. It performs safe string replacement without executing arbitrary code like f-strings would.

__init__

def __init__(prompts: dict[str, InterpolationPrompt])

Initialize the interpolation prompt engine.

Arguments:

• prompts: Map of prompt names to InterpolationPrompt objects

from_config

@classmethod
def from_config(cls, config: DictConfig) -> Self

Create InterpolationPromptEngine from configuration.

Expected config format:

{
    "type": "interpolation",
    "prompts": {
        "prompt_name": {
            "text": "Prompt string with {arg1} and {arg2}",
            "defaults": {
                "arg2": "default_value"
            }
        },
        "another_prompt": {
            "text": "Hello {name}, welcome to {service}!",
            "defaults": {
                "service": "our service"
            }
        }
    }
}

Arguments:

• config: Configuration data with prompts

Raises:

• ValueError: If prompts are missing from config

Returns:

InterpolationPromptEngine instance

generate

async def generate(target_name: str,
                   prompt: Prompt,
                   arguments: dict[str, str] | None = None) -> GetPromptResult

Generate a prompt response using prompt interpolation.

Safely substitutes argument values into the prompt string using simple string replacement without executing code. Uses default values for missing arguments when available.

Arguments:

• target_name: Name of the target that would handle the prompt call
• prompt: The MCP prompt definition
• arguments: Arguments to substitute into the prompt

Raises:

• ValueError: If prompt is not found or substitution fails

Returns:

Generated prompt result with interpolated content