Struct McpCreateMessageRequestParams 

#[non_exhaustive]
pub struct McpCreateMessageRequestParams {
    pub meta: Option<Meta>,
    pub task: Option<Map<String, Value>>,
    pub messages: Vec<SamplingMessage>,
    pub model_preferences: Option<ModelPreferences>,
    pub system_prompt: Option<String>,
    pub include_context: Option<ContextInclusion>,
    pub temperature: Option<f32>,
    pub max_tokens: u32,
    pub stop_sequences: Option<Vec<String>>,
    pub metadata: Option<Value>,
    pub tools: Option<Vec<Tool>>,
    pub tool_choice: Option<ToolChoice>,
}

Re-exports of the rmcp wire-protocol types this crate now surfaces directly instead of wrapping. Pull these in to pattern-match on tool annotations, content blocks, structured tool output, embedded resources, sampling / elicitation requests, progress and log notifications, etc. Parameters for creating a message through LLM sampling.

This structure contains all the necessary information for a client to generate an LLM response, including conversation history, model preferences, and generation parameters.

This implements TaskAugmentedRequestParamsMeta as sampling requests can be long-running and may benefit from task-based execution.

Fields (Non-exhaustive)

Non-exhaustive structs could have additional fields added in future. Therefore, non-exhaustive structs cannot be constructed in external crates using the traditional Struct { .. } syntax; cannot be matched against without a wildcard ..; and struct update syntax will not work.

meta: Option<Meta>

Protocol-level metadata for this request (SEP-1319)

task: Option<Map<String, Value>>

Task metadata for async task management (SEP-1319)

messages: Vec<SamplingMessage>

The conversation history and current messages

model_preferences: Option<ModelPreferences>

Preferences for model selection and behavior

system_prompt: Option<String>

System prompt to guide the model’s behavior

include_context: Option<ContextInclusion>

How much context to include from MCP servers

temperature: Option<f32>

Temperature for controlling randomness (0.0 to 1.0)

max_tokens: u32

Maximum number of tokens to generate

stop_sequences: Option<Vec<String>>

Sequences that should stop generation

metadata: Option<Value>

Additional metadata for the request

tools: Option<Vec<Tool>>

Tools available for the model to call (SEP-1577)

tool_choice: Option<ToolChoice>

Tool selection behavior (SEP-1577)

Implementations

impl CreateMessageRequestParams

pub fn new( messages: Vec<SamplingMessage>, max_tokens: u32, ) -> CreateMessageRequestParams

Create a new CreateMessageRequestParams with required fields.

pub fn with_model_preferences( self, model_preferences: ModelPreferences, ) -> CreateMessageRequestParams

Set model preferences.

pub fn with_system_prompt( self, system_prompt: impl Into<String>, ) -> CreateMessageRequestParams

Set system prompt.

pub fn with_include_context( self, include_context: ContextInclusion, ) -> CreateMessageRequestParams

Set include context.

pub fn with_temperature(self, temperature: f32) -> CreateMessageRequestParams

Set temperature.

pub fn with_stop_sequences( self, stop_sequences: Vec<String>, ) -> CreateMessageRequestParams

Set stop sequences.

pub fn with_metadata(self, metadata: Value) -> CreateMessageRequestParams

Set metadata.

pub fn with_tools(self, tools: Vec<Tool>) -> CreateMessageRequestParams

Set tools.

pub fn with_tool_choice( self, tool_choice: ToolChoice, ) -> CreateMessageRequestParams

Set tool choice.

pub fn validate(&self) -> Result<(), String>

Validate the sampling request parameters per SEP-1577 spec requirements.

Checks:

• ToolUse content is only allowed in assistant messages
• ToolResult content is only allowed in user messages
• Messages with tool result content MUST NOT contain other content types
• Every assistant ToolUse must be balanced with a corresponding user ToolResult

Trait Implementations

impl Clone for CreateMessageRequestParams

fn clone(&self) -> CreateMessageRequestParams

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for CreateMessageRequestParams

fn fmt(&self, f: &mut Formatter<'_>) -> Result<(), Error>

Formats the value using the given formatter.

impl Default for CreateMessageRequestParams

fn default() -> CreateMessageRequestParams

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for CreateMessageRequestParams

fn deserialize<__D>( __deserializer: __D, ) -> Result<CreateMessageRequestParams, <__D as Deserializer<'de>>::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl PartialEq for CreateMessageRequestParams

fn eq(&self, other: &CreateMessageRequestParams) -> bool

Tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl RequestParamsMeta for CreateMessageRequestParams

fn meta(&self) -> Option<&Meta>

Get a reference to the meta field

fn meta_mut(&mut self) -> &mut Option<Meta>

Get a mutable reference to the meta field

fn set_meta(&mut self, meta: Meta)

Set the meta field

fn progress_token(&self) -> Option<ProgressToken>

Get the progress token from meta, if present

fn set_progress_token(&mut self, token: ProgressToken)

Set a progress token in meta

impl Serialize for CreateMessageRequestParams

fn serialize<__S>( &self, __serializer: __S, ) -> Result<<__S as Serializer>::Ok, <__S as Serializer>::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.

impl StructuralPartialEq for CreateMessageRequestParams

impl TaskAugmentedRequestParamsMeta for CreateMessageRequestParams

fn task(&self) -> Option<&Map<String, Value>>

Get a reference to the task field

fn task_mut(&mut self) -> &mut Option<Map<String, Value>>

Get a mutable reference to the task field

fn set_task(&mut self, task: Map<String, Value>)

Set the task field