Enum AgentStreamEvent 

#[non_exhaustive]
pub enum AgentStreamEvent {
    Text(String),
    ToolStart {
        tool: String,
        input: String,
    },
    ToolEnd {
        tool: String,
        output: String,
    },
    Usage(TokenUsage),
    Finish {
        output: String,
    },
}

What an agent run is doing, streamed by AgentExecutor::stream_events.

Variants (Non-exhaustive)

Non-exhaustive enums could have additional variants added in future. Therefore, when matching against variants of non-exhaustive enums, an extra wildcard arm must be added to account for any future variants.

Text(String)

A chunk of text from the model, as it is generated. With tool-calling agents this is the answer itself (or a short preamble before tool calls).

ToolStart

The agent is about to run a tool.

Fields

tool: String

input: String

ToolEnd

A tool finished. output is what the agent sees; a failed tool shows up here as its error message unless the executor uses with_break_if_error(true).

Fields

tool: String

output: String

Usage(TokenUsage)

Tokens used by one model call of the run, when the LLM reports them. The run’s total is in Chain::call’s GenerateResult::tokens.

Finish

The run is over. Always the last event of a successful run.

Fields

output: String

Trait Implementations

impl Clone for AgentStreamEvent

fn clone(&self) -> Self

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.

impl PartialEq for AgentStreamEvent

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

Equality operator ==.

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

Inequality operator !=.

impl StructuralPartialEq for AgentStreamEvent