Enum Changed 

#[non_exhaustive]
pub enum Changed {
    Nothing,
    Message(MessageChange),
    MessagesReplaced,
    State,
    Reasoning(ReasoningChange),
    Subagent(SubagentChange),
    RunStarted {
        thread_id: ThreadId,
        run_id: RunId,
    },
    RunFinished {
        outcome: RunOutcome,
        result: Option<Value>,
    },
    RunError {
        message: String,
        code: Option<String>,
    },
}

What one event changed.

Returned by Applier::apply so a UI can redraw one row instead of the whole conversation. Every variant that names a message carries its index into Applier::messages.

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.

Nothing

Nothing a view would redraw: STEP_*, RAW, CUSTOM, or an event whose target does not exist.

Message(MessageChange)

One message was created, appended to, or completed.

MessagesReplaced

MESSAGES_SNAPSHOT replaced the whole list. Messages may have been removed, so a view must redraw all of it.

State

The application state was replaced or patched.

Reasoning(ReasoningChange)

Reasoning content changed. Reasoning is kept out of Applier::messages; see Applier::reasoning.

Subagent(SubagentChange)

A subagent was announced, resumed, finished, suspended or failed. What it produces arrives as ordinary message and reasoning changes, carrying its id; this is the lifecycle. See Applier::subagents.

RunStarted

RUN_STARTED.

Fields

thread_id: ThreadId

The conversation the run belongs to.

run_id: RunId

The run that started.

RunFinished

RUN_FINISHED. An absent outcome is reported as RunOutcome::Success, which is what the protocol says it means.

Fields

outcome: RunOutcome

How the run ended.

result: Option<Value>

The agent’s return value, if it sent one.

RunError

RUN_ERROR.

Fields

message: String

What went wrong, for a human.

code: Option<String>

The machine-readable code, when the agent sent one.

Trait Implementations

impl Clone for Changed

fn clone(&self) -> Changed

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for Changed

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

Formats the value using the given formatter.

impl PartialEq for Changed

fn eq(&self, other: &Changed) -> 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 StructuralPartialEq for Changed