mixle.telemetry.core module¶
Event schema and local-first recorder for platform telemetry.
An Event is a typed, timestamped decision record containing the
decision kind, the features used, the selected choice, and an optional outcome.
A Telemetry recorder buffers events and can append them to a JSONL log
for dashboards or learned policies.
- class Event(kind, features=<factory>, choice=None, outcome=<factory>, tags=<factory>, ts=0.0)[source]
Bases:
objectOne typed, timestamped decision record. Features/choice describe the decision; outcome scores it.
- Parameters:
- kind: str
- choice: Any = None
- ts: float = 0.0
- class Telemetry(path=None, *, flush_every=1)[source]
Bases:
objectA local-first event recorder: buffer in memory, append to a JSONL log, read back for training.
record(kind, features=..., choice=..., outcome=..., tags=...)appends oneEvent.events(kind=...)yields the buffer (optionally filtered).training_rows(kind)yields the(features, choice, outcome)triples the learned-orchestration models consume.- record(kind, *, features=None, choice=None, outcome=None, tags=None, when=None)[source]
Record one decision event; returns it (mutate
.outcomelater to close the loop).
- training_rows(kind)[source]
The
(features, choice, outcome)triples for a decision kind.
- flush()[source]
- Return type:
None
- get_default_recorder()[source]
The process-global recorder (a no-path in-memory buffer until one is configured).
- Return type:
Telemetry
- set_default_recorder(recorder)[source]
Install (or clear) the process-global recorder – e.g. point it at the user’s telemetry log.
- Parameters:
recorder (Telemetry | None)
- Return type:
None