core/chat/telemetry.py

"""ADR-0040 — structured-logging sink for turn-event audit.

Consumes ``TurnEvent`` records that ADR-0039 makes uniform across
main and stub paths.  Emits one JSON-line per turn with deterministic
field ordering, suitable for log aggregation, replay, and offline
audit pipelines.

Trust boundary (per CLAUDE.md):

* **Metadata-only by default.**  Surface text and input tokens are
  redacted unless the caller explicitly opts in via
  ``include_content=True``.  Audit needs counts, ids, and flags —
  not raw content — and the redact-by-default stance prevents
  accidental PII leakage when sinks point at shared log stores.
* **No implicit wall-clock.**  Timestamps are caller-provided so
  emission is reproducible under replay.  The runtime never reaches
  for ``datetime.now()`` here.
* **Append-only file paths.**  ``JsonlFileSink`` opens the target in
  append mode and never truncates.  Path is fixed at construction;
  the sink does not interpret user-controlled paths at emit time.
* **Idempotent flush.**  Each ``emit()`` flushes immediately so a
  crashed turn loop still has its prior turns durable on disk.

See ``docs/decisions/ADR-0040-telemetry-sink.md``.
"""

from __future__ import annotations

import json
from dataclasses import dataclass, field
from pathlib import Path
from typing import IO, Protocol


_UNKNOWN_DOMAIN_SURFACE = "I don't know — insufficient grounding for that yet."


# ---------- pure serializer ----------


def serialize_turn_event(
    event,
    *,
    safety_pack_id: str = "",
    ethics_pack_id: str = "",
    identity_pack_id: str = "",
    include_content: bool = False,
    timestamp: str | None = None,
) -> dict[str, object]:
    """Produce a JSON-safe audit dict from a ``TurnEvent``.

    Pack ids are passed as kwargs because ``TurnEvent`` does not
    carry them — the runtime knows them.  Fields are typed
    deliberately at the boundary so an upstream change to
    ``TurnEvent`` doesn't silently break the wire format; missing or
    differently-typed values fall back to safe defaults.
    """
    verdicts = getattr(event, "verdicts", None)
    out: dict[str, object] = {
        "turn": int(getattr(event, "turn", 0)),
        "safety_pack_id": str(safety_pack_id),
        "ethics_pack_id": str(ethics_pack_id),
        "identity_pack_id": str(identity_pack_id),
        "refusal_emitted": bool(getattr(verdicts, "refusal_emitted", False)),
        "hedge_injected": bool(getattr(verdicts, "hedge_injected", False)),
        "versor_condition": float(getattr(event, "versor_condition", 0.0)),
        "vault_hits": int(getattr(event, "vault_hits", 0)),
        "cycle_cost_total": float(getattr(event, "cycle_cost_total", 0.0)),
        "flagged": bool(getattr(event, "flagged", False)),
        "stub_path": getattr(event, "walk_surface", "") == _UNKNOWN_DOMAIN_SURFACE,
        "dialogue_role": str(getattr(event, "dialogue_role", "")),
    }
    safety = getattr(event, "safety_verdict", None)
    if safety is not None:
        out["safety_violated"] = sorted(
            getattr(safety, "violated_boundaries", ()) or ()
        )
        out["safety_runtime_checkable_count"] = int(
            getattr(safety, "runtime_checkable_count", 0)
        )
        out["safety_upheld"] = bool(getattr(safety, "upheld", True))
    ethics = getattr(event, "ethics_verdict", None)
    if ethics is not None:
        out["ethics_violated"] = sorted(
            getattr(ethics, "violated_commitments", ()) or ()
        )
        out["ethics_runtime_checkable_count"] = int(
            getattr(ethics, "runtime_checkable_count", 0)
        )
        out["ethics_upheld"] = bool(getattr(ethics, "upheld", True))
    identity_score = getattr(event, "identity_score", None)
    if identity_score is not None:
        out["identity_alignment"] = float(
            getattr(identity_score, "alignment", 0.0)
        )
        out["identity_flagged"] = bool(getattr(identity_score, "flagged", False))
        out["identity_deviation_axes"] = sorted(
            getattr(identity_score, "deviation_axes", ()) or ()
        )
    if include_content:
        out["input_tokens"] = list(getattr(event, "input_tokens", ()))
        out["surface"] = str(getattr(event, "surface", ""))
        out["walk_surface"] = str(getattr(event, "walk_surface", ""))
        out["articulation_surface"] = str(
            getattr(event, "articulation_surface", "")
        )
    if timestamp is not None:
        out["timestamp"] = str(timestamp)
    return out


def format_turn_event_jsonl(event, **kwargs) -> str:
    """Serialize a turn event as one deterministic JSONL line.

    Field order is alphabetical (``sort_keys=True``) so two emissions
    of the same logical event produce byte-identical lines.  No
    trailing newline — the sink owns line termination.
    """
    payload = serialize_turn_event(event, **kwargs)
    return json.dumps(payload, sort_keys=True, separators=(",", ":"))


# ---------- sink protocol ----------


class TurnEventSink(Protocol):
    """Minimal sink contract.

    Sinks receive one already-serialized JSONL line per turn.  The
    runtime calls ``emit()`` after each ``turn_log.append()`` — see
    ``ChatRuntime._emit_turn_event``.
    """

    def emit(self, line: str) -> None: ...


# ---------- concrete sinks ----------


@dataclass
class JsonlBufferSink:
    """In-memory sink that captures every emitted line.

    Useful for tests, replay diffing, and small-volume audit where
    persistence is the caller's responsibility.
    """

    lines: list[str] = field(default_factory=list)

    def emit(self, line: str) -> None:
        self.lines.append(line)


class JsonlFileSink:
    """Append-only JSONL file sink with eager flush.

    The path is fixed at construction.  Each ``emit()`` flushes
    immediately so a crashed runtime still has its prior turns
    durable on disk.  Supports context-manager usage.
    """

    def __init__(self, path: str | Path) -> None:
        self._path = Path(path)
        self._fh: IO[str] | None = None

    def emit(self, line: str) -> None:
        if self._fh is None:
            self._path.parent.mkdir(parents=True, exist_ok=True)
            self._fh = self._path.open("a", encoding="utf-8")
        self._fh.write(line)
        self._fh.write("\n")
        self._fh.flush()

    def close(self) -> None:
        if self._fh is not None:
            self._fh.close()
            self._fh = None

    def __enter__(self) -> "JsonlFileSink":
        return self

    def __exit__(self, *exc_info) -> None:
        self.close()


__all__ = [
    "JsonlBufferSink",
    "JsonlFileSink",
    "TurnEventSink",
    "format_turn_event_jsonl",
    "serialize_turn_event",
]