pub struct OutcomeRecord {
pub id: String,
pub task_id: Option<String>,
pub tool_run_id: Option<String>,
pub kind: String,
pub label: String,
pub reward: Option<f64>,
pub source: String,
pub detail_json: Option<String>,
pub created_at: String,
}Expand description
A verifiable outcome / reward signal attached to a trajectory (a task, and
optionally a specific tool run). The other durable tables record what
happened (messages, tool_runs, checkpoints=diffs); outcomes records how
good it was and who says so (source) — the enrichment
that turns logs into a training set for the self-improving loop.
Fields§
§id: String§task_id: Option<String>§tool_run_id: Option<String>§kind: StringSignal type, e.g. task_terminal, build, test, tool_exec,
user_edit, git_survival, preference. Free-form.
label: StringGraded result — one of the OUTCOME_LABEL_* values.
reward: Option<f64>Optional scalar reward (convention: roughly -1.0..=1.0). None when
the signal is categorical only.
source: StringProvenance — one of the OUTCOME_SOURCE_* values.
detail_json: Option<String>Optional structured payload: test counts, a git sha, or a preference
pair { "chosen": ..., "rejected": ... } for DPO.
created_at: StringTrait Implementations§
Source§impl Clone for OutcomeRecord
impl Clone for OutcomeRecord
Source§fn clone(&self) -> OutcomeRecord
fn clone(&self) -> OutcomeRecord
1.0.0 (const: unstable) · Source§fn clone_from(&mut self, source: &Self)
fn clone_from(&mut self, source: &Self)
source. Read more