Enum AgentStreamEvent 

pub enum AgentStreamEvent {
    RunStart {
        run_id: String,
    },
    ContextInfo {
        estimated_tokens: usize,
        request_bytes: usize,
        context_limit: Option<u64>,
    },
    ContextCompressed {
        original_tokens: usize,
        compressed_tokens: usize,
        strategy: String,
        messages_before: usize,
        messages_after: usize,
    },
    RequestStart {
        step: u32,
    },
    TextDelta {
        text: String,
    },
    ToolCallStart {
        tool_name: String,
        tool_call_id: Option<String>,
    },
    ToolCallDelta {
        delta: String,
        tool_call_id: Option<String>,
    },
    ToolCallComplete {
        tool_name: String,
        tool_call_id: Option<String>,
    },
    ToolExecuted {
        tool_name: String,
        tool_call_id: Option<String>,
        success: bool,
        error: Option<String>,
    },
    ThinkingDelta {
        text: String,
    },
    ResponseComplete {
        step: u32,
    },
    OutputReady,
    RunComplete {
        run_id: String,
    },
    Error {
        message: String,
    },
    Cancelled {
        partial_text: Option<String>,
        partial_thinking: Option<String>,
        pending_tools: Vec<String>,
    },
}

Events emitted during streaming.

Variants

RunStart

Run started.

Fields

run_id: String

ContextInfo

Context size information (emitted before each model request).

Fields

estimated_tokens: usize

Estimated token count (~request_bytes / 4).

request_bytes: usize

Raw request size in bytes (serialized messages + tools).

context_limit: Option<u64>

Model’s context window limit (if known).

ContextCompressed

Context was compressed to fit within limits.

Fields

original_tokens: usize

Token count before compression.

compressed_tokens: usize

Token count after compression.

strategy: String

Strategy used: “truncate” or “summarize”.

messages_before: usize

Number of messages before compression.

messages_after: usize

Number of messages after compression.

RequestStart

Model request started.

Fields

step: u32

TextDelta

Text delta.

Fields

text: String

ToolCallStart

Tool call started.

Fields

tool_name: String

tool_call_id: Option<String>

ToolCallDelta

Tool call arguments delta.

Fields

delta: String

tool_call_id: Option<String>

ToolCallComplete

Tool call completed (arguments fully received).

Fields

tool_name: String

tool_call_id: Option<String>

ToolExecuted

Tool executed.

Fields

tool_name: String

tool_call_id: Option<String>

success: bool

error: Option<String>

ThinkingDelta

Thinking delta (for reasoning models).

Fields

text: String

ResponseComplete

Model response completed.

Fields

step: u32

OutputReady

Output ready.

RunComplete

Run completed.

Fields

run_id: String

Error

Error occurred.

Fields

message: String

Cancelled

Run was cancelled.

Fields

partial_text: Option<String>

Partial text accumulated before cancellation.

partial_thinking: Option<String>

Partial thinking content accumulated before cancellation.

pending_tools: Vec<String>

Tool calls that were in progress when cancelled.

Trait Implementations

impl Clone for AgentStreamEvent

fn clone(&self) -> AgentStreamEvent

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for AgentStreamEvent

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

Formats the value using the given formatter.