Enum ChatEvent 

pub enum ChatEvent {
    SessionStarted {
        meta: EventMeta,
        session_id: String,
        model: String,
    },
    SessionCompleted {
        meta: EventMeta,
        session_id: String,
        status: SessionStatus,
    },
    RoundStarted {
        meta: EventMeta,
        round_id: String,
        session_id: String,
        model: String,
    },
    RoundCompleted {
        meta: EventMeta,
        round_id: String,
        session_id: String,
        status: RoundStatus,
        usage: TokenUsage,
        latency_ms: u64,
        error: Option<String>,
    },
    ToolCalled {
        meta: EventMeta,
        tool_call_id: String,
        round_id: String,
        session_id: String,
        tool_name: String,
        latency_ms: u64,
        success: bool,
    },
    MessageCountUpdated {
        meta: EventMeta,
        session_id: String,
        message_count: u32,
    },
}

Events emitted by the agent loop during chat sessions.

These events track the complete lifecycle of chat interactions within the agent, from session initialization to completion, including all intermediate operations like rounds and tool calls.

Event Lifecycle

A typical chat session produces events in this order:

1. SessionStarted - When a new conversation begins
2. RoundStarted - When the agent begins processing a message
3. ToolCalled (multiple) - For each tool invocation during the round
4. RoundCompleted - When the agent finishes processing the message
5. MessageCountUpdated - When the message count changes
6. SessionCompleted - When the conversation ends

Usage

These events are primarily used for:

• Performance monitoring (latency tracking)
• Resource usage analysis (token consumption)
• Behavior analytics (tool usage patterns)
• Error tracking and debugging

Variants

SessionStarted

Emitted when a new chat session is initialized.

This marks the beginning of a conversation between the user and agent.

Fields

meta: EventMeta

Event metadata including unique ID and timestamp

session_id: String

Unique identifier for this chat session

model: String

The AI model being used for this session (e.g., “gpt-4”, “claude-3”)

SessionCompleted

Emitted when a chat session completes or terminates.

This marks the end of a conversation, whether successful or due to an error.

Fields

meta: EventMeta

Event metadata including unique ID and timestamp

session_id: String

Unique identifier for the chat session that completed

status: SessionStatus

Final status of the session (completed, failed, or cancelled)

RoundStarted

Emitted when the agent starts processing a new message round.

A round represents a single request-response cycle within a session. Each user message typically triggers one round.

Fields

meta: EventMeta

Event metadata including unique ID and timestamp

round_id: String

Unique identifier for this round

session_id: String

Session this round belongs to

model: String

The AI model being used for this round

RoundCompleted

Emitted when the agent completes processing a message round.

Contains comprehensive metrics about the round including token usage, latency, and any errors that occurred.

Fields

meta: EventMeta

Event metadata including unique ID and timestamp

round_id: String

Unique identifier for the round that completed

session_id: String

Session this round belongs to

status: RoundStatus

Final status of the round (success or failed)

usage: TokenUsage

Token consumption during this round

latency_ms: u64

Total time to process the round in milliseconds

error: Option<String>

Error message if the round failed, None on success

ToolCalled

Emitted when the agent invokes a tool during a round.

Tracks tool execution for understanding agent behavior and measuring tool performance.

Fields

meta: EventMeta

Event metadata including unique ID and timestamp

tool_call_id: String

Unique identifier for this tool invocation

round_id: String

Round this tool call belongs to

session_id: String

Session this tool call belongs to

tool_name: String

Name of the tool being invoked (e.g., “read_file”, “execute_command”)

latency_ms: u64

Time taken to execute the tool in milliseconds

success: bool

Whether the tool execution succeeded

MessageCountUpdated

Emitted when the message count for a session is updated.

This tracks the total number of messages exchanged in the session, including both user and assistant messages.

Fields

meta: EventMeta

Event metadata including unique ID and timestamp

session_id: String

Session whose message count was updated

message_count: u32

New total message count for the session

Trait Implementations

impl Clone for ChatEvent

fn clone(&self) -> ChatEvent

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for ChatEvent

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

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for ChatEvent

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for ChatEvent

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

where __S: Serializer,

Serialize this value into the given Serde serializer.