Struct ExecutorEvent 

pub struct ExecutorEvent {
    pub execution_id: i64,
    pub event_type: String,
    pub step: String,
    pub status: String,
    pub created_at: DateTime<Utc>,
    pub context: Value,
    pub event_id: Option<i64>,
    pub worker_id: Option<String>,
    pub meta: Option<Value>,
}

One event the executor emits. Keep field naming aligned with the Python side (noetl.event table columns) so envelopes can be projected by either stack.

R-1.2 PR-2a: execution_id is now i64 (matching the Python noetl.event.execution_id bigint column, the CLI’s BridgeContext.execution_id, the worker’s CommandNotification.execution_id, and noetl_tools::context::ExecutionContext.execution_id).

R-1.2 PR-EE-1 (0.3.1): adds the optional fields the Python EventEmitRequest (and the Rust noetl-server EventRequest) already expect — event_id, worker_id, meta. All are Option + #[serde(default, skip_serializing_if = "Option::is_none")] so older wire-format consumers deserialize cleanly and producers that don’t populate them omit them from the JSON entirely. This step is the additive prerequisite for the cross-repo event envelope reconciliation tracked on noetl/ai-meta#30.

EE-4 (noetl/ai-meta#49): the struct moved out of noetl-executor::events into this dedicated noetl-events crate so noetl-server can link it directly.

Fields

execution_id: i64

Execution this event belongs to. Matches the noetl.event.execution_id bigint column.

event_type: String

Event type (e.g. "step.enter", "call.done", "step.exit", "command.completed", "command.failed"). Python side uses an EventType enum but Rust accepts the raw string; the server validates against the enum when projecting.

step: String

Step name (node_id / node_name in the Python projector).

status: String

Lifecycle status (e.g. "STARTED", "COMPLETED", "FAILED").

created_at: DateTime<Utc>

Wall-clock when the event was produced. Stamped at emit time so the event log preserves per-component ordering even across server-clock skew.

context: Value

Free-form payload; the projector reads typed fields out of this. Renamed from the worker’s payload field to match the Python EventEmitRequest.context field; serde alias accepts either name on the wire.

event_id: Option<i64>

Application-side snowflake id for this event. Per agents/rules/observability.md Principle 3, the emitting process generates this BEFORE the row hits the database so spans / metrics / cross-component correlation can use it immediately. None falls back to the DB-side default (existing gen_snowflake() function).

worker_id: Option<String>

Worker that emitted the event (worker pod’s id, or "cli-local" for CLI mode). Used for shard-aware queries

• diagnostic correlation.

meta: Option<Value>

Free-form metadata that doesn’t belong in context — retry counts, parent-event refs, catalog ids. Matches the Python EventEmitRequest.meta field 1:1.

Trait Implementations

impl Clone for ExecutorEvent

fn clone(&self) -> ExecutorEvent

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for ExecutorEvent

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

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for ExecutorEvent

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

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for ExecutorEvent

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.