File size: 28,746 Bytes
ebcde1f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
"""Exact, outcome-blind contract for the frozen M8 live-L2 analysis.

The capture configuration controls what is observed.  This independent file
controls how already-verified session bundles may be analysed.  The authority
loader requires both the frozen semantics and the exact reviewed TOML bytes;
the semantic-hash helper exists only for provenance comparisons and does not
authorize a differently encoded configuration.
"""

from __future__ import annotations

import hashlib
import json
import math
import tomllib
from collections.abc import Mapping
from dataclasses import dataclass
from pathlib import Path
from typing import Any, Literal, cast

M8L2AnalysisRole = Literal["train", "validation", "primary_test", "replication_test"]
M8L2EndpointDomain = Literal["event", "clock"]
M8L2EndpointUnit = Literal["events", "milliseconds"]

M8_L2_ANALYSIS_CONFIG_SOURCE_SHA256 = (
    "0d786d5f4109bb5bf773a6197df3fa861c9b7eb61c16c957bd49fb56147fd7d8"
)
M8_L2_ANALYSIS_CONFIG_SEMANTIC_SHA256 = (
    "17c91f64765f35195ab03a4caac93d8ff9c5f009c16e785fd84ebd9569d6f84b"
)

_TOP_LEVEL_KEYS = frozenset(
    {
        "study",
        "features",
        "endpoints",
        "regimes",
        "calibration",
        "bootstrap",
        "signed_impact",
        "execution",
        "claims",
    }
)
_STUDY_KEYS = frozenset(
    {
        "name",
        "protocol_version",
        "seed",
        "source",
        "capture_config_source_sha256",
        "capture_protocol_sha256",
        "symbols",
        "training_role",
        "selection_role",
        "primary_endpoint_role",
        "replication_endpoint_role",
    }
)
_FEATURE_KEYS = frozenset(
    {
        "decision_scope",
        "flat_direction_policy",
        "rolling_windows",
        "volatility_window",
        "model_feature_columns",
        "clock_max_state_age_ms",
        "clock_target_policy",
        "clock_label_information_end",
        "clock_record_target_sequence",
        "clock_censor_if_no_eligible_state",
    }
)
_ENDPOINT_KEYS = frozenset(
    {
        "name",
        "domain",
        "horizon_value",
        "unit",
        "paired_block_width",
        "paired_block_unit",
        "nominal_event_block_width",
    }
)
_REGIME_KEYS = frozenset({"fit_role", "feature", "quantile_numerators", "quantile_denominator"})
_CALIBRATION_KEYS = frozenset({"bins"})
_BOOTSTRAP_KEYS = frozenset({"method", "samples"})
_SIGNED_IMPACT_KEYS = frozenset({"metric", "side_rule", "price_rule"})
_EXECUTION_KEYS = frozenset(
    {
        "market_orders_only",
        "probability_threshold",
        "symmetric_probability_thresholds",
        "order_notional_usd",
        "max_l1_participation",
        "inventory_order_multiples",
        "reference_price_fit_role",
        "reference_depth_fit_role",
        "reference_price_statistic",
        "reference_depth_statistic",
        "reference_quantity_policy",
        "l1_fill_policy",
        "scenario_reset_policy",
        "extra_slippage_bps",
        "liquidate_at_end",
    }
)
_CLAIM_KEYS = frozenset(
    {
        "allow_capacity_claim",
        "allow_realized_execution_claim",
        "allow_profitability_claim",
    }
)

_CAPTURE_CONFIG_SOURCE_SHA256 = "b1bf3b4e2820e24e4555bfeb9cb0957f9a0bcdef62039f7d92360e0a97d0dd39"
_CAPTURE_PROTOCOL_SHA256 = "4c77a2099a4cabd049d10e0f8264d3b4c66704d8e87cbaf0c817fd085f4bbd83"
_MODEL_FEATURE_COLUMNS = (
    "spread_bps",
    "depth_total_l1",
    "depth_total_l5",
    "depth_total_l10",
    "queue_imbalance_l1",
    "queue_imbalance_l5",
    "queue_imbalance_l10",
    "microprice_deviation_bps",
    "ofi_l1",
    "ofi_w20",
    "ofi_w100",
    "cancellation_intensity_w20",
    "cancellation_intensity_w100",
    "realized_volatility_w20",
    "realized_volatility_w100",
    "volatility_regime_low",
    "volatility_regime_high",
    "liquidity_regime_liquid",
    "liquidity_regime_stressed",
)


class M8L2AnalysisConfigError(ValueError):
    """Raised when an analysis file differs from the frozen contract."""


@dataclass(frozen=True, slots=True)
class M8L2AnalysisStudy:
    name: str
    protocol_version: str
    seed: int
    source: str
    capture_config_source_sha256: str
    capture_protocol_sha256: str
    symbols: tuple[str, ...]
    training_role: M8L2AnalysisRole
    selection_role: M8L2AnalysisRole
    primary_endpoint_role: M8L2AnalysisRole
    replication_endpoint_role: M8L2AnalysisRole


@dataclass(frozen=True, slots=True)
class M8L2AnalysisFeatures:
    decision_scope: str
    flat_direction_policy: str
    rolling_windows: tuple[int, ...]
    volatility_window: int
    model_feature_columns: tuple[str, ...]
    clock_max_state_age_ms: int
    clock_target_policy: str
    clock_label_information_end: str
    clock_record_target_sequence: bool
    clock_censor_if_no_eligible_state: bool


@dataclass(frozen=True, slots=True)
class M8L2AnalysisEndpoint:
    name: str
    domain: M8L2EndpointDomain
    horizon_value: int
    unit: M8L2EndpointUnit
    paired_block_width: int
    paired_block_unit: M8L2EndpointUnit
    nominal_event_block_width: int


@dataclass(frozen=True, slots=True)
class M8L2AnalysisRegimes:
    fit_role: M8L2AnalysisRole
    feature: str
    quantile_numerators: tuple[int, ...]
    quantile_denominator: int


@dataclass(frozen=True, slots=True)
class M8L2AnalysisCalibration:
    bins: int


@dataclass(frozen=True, slots=True)
class M8L2AnalysisBootstrap:
    method: str
    samples: int


@dataclass(frozen=True, slots=True)
class M8L2AnalysisSignedImpact:
    metric: str
    side_rule: str
    price_rule: str


@dataclass(frozen=True, slots=True)
class M8L2AnalysisExecution:
    market_orders_only: bool
    probability_threshold: float
    symmetric_probability_thresholds: bool
    order_notional_usd: float
    max_l1_participation: float
    inventory_order_multiples: int
    reference_price_fit_role: M8L2AnalysisRole
    reference_depth_fit_role: M8L2AnalysisRole
    reference_price_statistic: str
    reference_depth_statistic: str
    reference_quantity_policy: str
    l1_fill_policy: str
    scenario_reset_policy: str
    extra_slippage_bps: float
    liquidate_at_end: bool


@dataclass(frozen=True, slots=True)
class M8L2AnalysisClaims:
    allow_capacity_claim: bool
    allow_realized_execution_claim: bool
    allow_profitability_claim: bool


@dataclass(frozen=True, slots=True)
class M8L2AnalysisConfig:
    """Typed analysis contract, with separate semantic and exact-byte identities."""

    path: Path
    source_sha256: str
    study: M8L2AnalysisStudy
    features: M8L2AnalysisFeatures
    endpoints: tuple[M8L2AnalysisEndpoint, ...]
    regimes: M8L2AnalysisRegimes
    calibration: M8L2AnalysisCalibration
    bootstrap: M8L2AnalysisBootstrap
    signed_impact: M8L2AnalysisSignedImpact
    execution: M8L2AnalysisExecution
    claims: M8L2AnalysisClaims

    def _semantic_payload(self) -> dict[str, object]:
        return {
            "study": {name: getattr(self.study, name) for name in self.study.__dataclass_fields__},
            "features": {
                "decision_scope": self.features.decision_scope,
                "flat_direction_policy": self.features.flat_direction_policy,
                "rolling_windows": list(self.features.rolling_windows),
                "volatility_window": self.features.volatility_window,
                "model_feature_columns": list(self.features.model_feature_columns),
                "clock_max_state_age_ms": self.features.clock_max_state_age_ms,
                "clock_target_policy": self.features.clock_target_policy,
                "clock_label_information_end": self.features.clock_label_information_end,
                "clock_record_target_sequence": self.features.clock_record_target_sequence,
                "clock_censor_if_no_eligible_state": (
                    self.features.clock_censor_if_no_eligible_state
                ),
            },
            "endpoints": [
                {name: getattr(endpoint, name) for name in endpoint.__dataclass_fields__}
                for endpoint in self.endpoints
            ],
            "regimes": {
                "fit_role": self.regimes.fit_role,
                "feature": self.regimes.feature,
                "quantile_numerators": list(self.regimes.quantile_numerators),
                "quantile_denominator": self.regimes.quantile_denominator,
            },
            "calibration": {"bins": self.calibration.bins},
            "bootstrap": {
                name: getattr(self.bootstrap, name) for name in self.bootstrap.__dataclass_fields__
            },
            "signed_impact": {
                "metric": self.signed_impact.metric,
                "side_rule": self.signed_impact.side_rule,
                "price_rule": self.signed_impact.price_rule,
            },
            "execution": {
                name: getattr(self.execution, name) for name in self.execution.__dataclass_fields__
            },
            "claims": {
                name: getattr(self.claims, name) for name in self.claims.__dataclass_fields__
            },
        }

    @property
    def semantic_sha256(self) -> str:
        encoded = json.dumps(
            self._semantic_payload(), sort_keys=True, separators=(",", ":"), allow_nan=False
        ).encode("utf-8")
        return hashlib.sha256(encoded).hexdigest()

    @property
    def hash(self) -> str:
        """Compatibility alias for the formatting-independent semantic identity."""

        return self.semantic_sha256

    def public_dict(self) -> dict[str, object]:
        return {
            "path": str(self.path),
            "config_sha256": self.semantic_sha256,
            "semantic_sha256": self.semantic_sha256,
            "source_sha256": self.source_sha256,
            **self._semantic_payload(),
        }


def _mapping(value: object, label: str) -> Mapping[str, Any]:
    if not isinstance(value, Mapping) or not all(type(key) is str for key in value):
        raise M8L2AnalysisConfigError(f"{label} must be a TOML table with string keys")
    return cast(Mapping[str, Any], value)


def _exact_keys(value: Mapping[str, Any], expected: frozenset[str], label: str) -> None:
    observed = frozenset(value)
    if observed != expected:
        raise M8L2AnalysisConfigError(
            f"{label} keys differ (missing={sorted(expected - observed)}, "
            f"unknown={sorted(observed - expected)})"
        )


def _list(value: object, label: str) -> list[Any]:
    if not isinstance(value, list):
        raise M8L2AnalysisConfigError(f"{label} must be an array")
    return value


def _text(value: object, label: str) -> str:
    if type(value) is not str:
        raise M8L2AnalysisConfigError(f"{label} must be a string")
    return value


def _integer(value: object, label: str) -> int:
    if type(value) is not int:
        raise M8L2AnalysisConfigError(f"{label} must be an integer")
    return value


def _number(value: object, label: str) -> float:
    if type(value) not in {int, float}:
        raise M8L2AnalysisConfigError(f"{label} must be a finite number")
    result = float(cast(int | float, value))
    if not math.isfinite(result):
        raise M8L2AnalysisConfigError(f"{label} must be a finite number")
    return result


def _boolean(value: object, label: str) -> bool:
    if type(value) is not bool:
        raise M8L2AnalysisConfigError(f"{label} must be a boolean")
    return value


def _text_tuple(value: object, label: str) -> tuple[str, ...]:
    return tuple(_text(item, f"{label}[{index}]") for index, item in enumerate(_list(value, label)))


def _integer_tuple(value: object, label: str) -> tuple[int, ...]:
    return tuple(
        _integer(item, f"{label}[{index}]") for index, item in enumerate(_list(value, label))
    )


def _frozen(observed: object, expected: object, label: str) -> None:
    if observed != expected:
        raise M8L2AnalysisConfigError(f"{label} is frozen at {expected!r}, observed {observed!r}")


def _role(value: object, label: str) -> M8L2AnalysisRole:
    observed = _text(value, label)
    if observed not in {"train", "validation", "primary_test", "replication_test"}:
        raise M8L2AnalysisConfigError(f"{label} is not a supported frozen-session role")
    return cast(M8L2AnalysisRole, observed)


def _parse_study(raw: object) -> M8L2AnalysisStudy:
    table = _mapping(raw, "study")
    _exact_keys(table, _STUDY_KEYS, "study")
    result = M8L2AnalysisStudy(
        name=_text(table["name"], "study.name"),
        protocol_version=_text(table["protocol_version"], "study.protocol_version"),
        seed=_integer(table["seed"], "study.seed"),
        source=_text(table["source"], "study.source"),
        capture_config_source_sha256=_text(
            table["capture_config_source_sha256"], "study.capture_config_source_sha256"
        ),
        capture_protocol_sha256=_text(
            table["capture_protocol_sha256"], "study.capture_protocol_sha256"
        ),
        symbols=_text_tuple(table["symbols"], "study.symbols"),
        training_role=_role(table["training_role"], "study.training_role"),
        selection_role=_role(table["selection_role"], "study.selection_role"),
        primary_endpoint_role=_role(table["primary_endpoint_role"], "study.primary_endpoint_role"),
        replication_endpoint_role=_role(
            table["replication_endpoint_role"], "study.replication_endpoint_role"
        ),
    )
    expected = M8L2AnalysisStudy(
        name="binance-m8-live-l2-analysis-v2",
        protocol_version="2.0.0",
        seed=20260807,
        source="verified_m8_l2_session_bundles",
        capture_config_source_sha256=_CAPTURE_CONFIG_SOURCE_SHA256,
        capture_protocol_sha256=_CAPTURE_PROTOCOL_SHA256,
        symbols=("BTCUSDT", "ETHUSDT"),
        training_role="train",
        selection_role="validation",
        primary_endpoint_role="primary_test",
        replication_endpoint_role="replication_test",
    )
    _frozen(result, expected, "study contract")
    return result


def _parse_features(raw: object) -> M8L2AnalysisFeatures:
    table = _mapping(raw, "features")
    _exact_keys(table, _FEATURE_KEYS, "features")
    result = M8L2AnalysisFeatures(
        decision_scope=_text(table["decision_scope"], "features.decision_scope"),
        flat_direction_policy=_text(
            table["flat_direction_policy"], "features.flat_direction_policy"
        ),
        rolling_windows=_integer_tuple(table["rolling_windows"], "features.rolling_windows"),
        volatility_window=_integer(table["volatility_window"], "features.volatility_window"),
        model_feature_columns=_text_tuple(
            table["model_feature_columns"], "features.model_feature_columns"
        ),
        clock_max_state_age_ms=_integer(
            table["clock_max_state_age_ms"],
            "features.clock_max_state_age_ms",
        ),
        clock_target_policy=_text(table["clock_target_policy"], "features.clock_target_policy"),
        clock_label_information_end=_text(
            table["clock_label_information_end"], "features.clock_label_information_end"
        ),
        clock_record_target_sequence=_boolean(
            table["clock_record_target_sequence"], "features.clock_record_target_sequence"
        ),
        clock_censor_if_no_eligible_state=_boolean(
            table["clock_censor_if_no_eligible_state"],
            "features.clock_censor_if_no_eligible_state",
        ),
    )
    if any(value <= 0 for value in (*result.rolling_windows, result.volatility_window)):
        raise M8L2AnalysisConfigError("feature windows must be positive")
    if result.clock_max_state_age_ms < 0:
        raise M8L2AnalysisConfigError("features.clock_max_state_age_ms must be nonnegative")
    _frozen(
        result,
        M8L2AnalysisFeatures(
            decision_scope="per_symbol_verified_observed_intervals",
            flat_direction_policy="flat_is_non_up",
            rolling_windows=(20, 100),
            volatility_window=100,
            model_feature_columns=_MODEL_FEATURE_COLUMNS,
            clock_max_state_age_ms=500,
            clock_target_policy="exact_target_locf_same_valid_observed_interval",
            clock_label_information_end="exact_target",
            clock_record_target_sequence=True,
            clock_censor_if_no_eligible_state=True,
        ),
        "features contract",
    )
    return result


def _parse_endpoints(raw: object) -> tuple[M8L2AnalysisEndpoint, ...]:
    result: list[M8L2AnalysisEndpoint] = []
    for index, item in enumerate(_list(raw, "endpoints")):
        table = _mapping(item, f"endpoints[{index}]")
        _exact_keys(table, _ENDPOINT_KEYS, f"endpoints[{index}]")
        raw_domain = _text(table["domain"], f"endpoints[{index}].domain")
        if raw_domain not in {"event", "clock"}:
            raise M8L2AnalysisConfigError(f"endpoints[{index}].domain is unsupported")
        raw_unit = _text(table["unit"], f"endpoints[{index}].unit")
        if raw_unit not in {"events", "milliseconds"}:
            raise M8L2AnalysisConfigError(f"endpoints[{index}].unit is unsupported")
        raw_block_unit = _text(table["paired_block_unit"], f"endpoints[{index}].paired_block_unit")
        if raw_block_unit not in {"events", "milliseconds"}:
            raise M8L2AnalysisConfigError(f"endpoints[{index}].paired_block_unit is unsupported")
        endpoint = M8L2AnalysisEndpoint(
            name=_text(table["name"], f"endpoints[{index}].name"),
            domain=cast(M8L2EndpointDomain, raw_domain),
            horizon_value=_integer(table["horizon_value"], f"endpoints[{index}].horizon_value"),
            unit=cast(M8L2EndpointUnit, raw_unit),
            paired_block_width=_integer(
                table["paired_block_width"], f"endpoints[{index}].paired_block_width"
            ),
            paired_block_unit=cast(M8L2EndpointUnit, raw_block_unit),
            nominal_event_block_width=_integer(
                table["nominal_event_block_width"],
                f"endpoints[{index}].nominal_event_block_width",
            ),
        )
        if (
            endpoint.horizon_value <= 0
            or endpoint.paired_block_width <= 0
            or endpoint.nominal_event_block_width <= 0
        ):
            raise M8L2AnalysisConfigError("endpoint horizons and block widths must be positive")
        expected_unit = "events" if endpoint.domain == "event" else "milliseconds"
        if endpoint.unit != expected_unit or endpoint.paired_block_unit != expected_unit:
            raise M8L2AnalysisConfigError(
                f"endpoints[{index}] units do not match its endpoint domain"
            )
        result.append(endpoint)
    expected = (
        M8L2AnalysisEndpoint("event_20", "event", 20, "events", 40, "events", 40),
        M8L2AnalysisEndpoint("event_100", "event", 100, "events", 200, "events", 200),
        M8L2AnalysisEndpoint(
            "clock_1000ms", "clock", 1000, "milliseconds", 2000, "milliseconds", 20
        ),
        M8L2AnalysisEndpoint(
            "clock_5000ms", "clock", 5000, "milliseconds", 10000, "milliseconds", 100
        ),
    )
    _frozen(tuple(result), expected, "endpoint order/contract")
    return tuple(result)


def _parse_regimes(raw: object) -> M8L2AnalysisRegimes:
    table = _mapping(raw, "regimes")
    _exact_keys(table, _REGIME_KEYS, "regimes")
    result = M8L2AnalysisRegimes(
        fit_role=_role(table["fit_role"], "regimes.fit_role"),
        feature=_text(table["feature"], "regimes.feature"),
        quantile_numerators=_integer_tuple(
            table["quantile_numerators"], "regimes.quantile_numerators"
        ),
        quantile_denominator=_integer(
            table["quantile_denominator"], "regimes.quantile_denominator"
        ),
    )
    if result.quantile_denominator <= 0 or any(
        value <= 0 or value >= result.quantile_denominator for value in result.quantile_numerators
    ):
        raise M8L2AnalysisConfigError("regime quantiles must lie strictly between zero and one")
    _frozen(
        result,
        M8L2AnalysisRegimes("train", "realized_volatility_w100", (1, 2), 3),
        "regimes contract",
    )
    return result


def _parse_calibration(raw: object) -> M8L2AnalysisCalibration:
    table = _mapping(raw, "calibration")
    _exact_keys(table, _CALIBRATION_KEYS, "calibration")
    result = M8L2AnalysisCalibration(bins=_integer(table["bins"], "calibration.bins"))
    if result.bins < 2:
        raise M8L2AnalysisConfigError("calibration.bins must be at least two")
    _frozen(result, M8L2AnalysisCalibration(10), "calibration contract")
    return result


def _parse_bootstrap(raw: object) -> M8L2AnalysisBootstrap:
    table = _mapping(raw, "bootstrap")
    _exact_keys(table, _BOOTSTRAP_KEYS, "bootstrap")
    result = M8L2AnalysisBootstrap(
        method=_text(table["method"], "bootstrap.method"),
        samples=_integer(table["samples"], "bootstrap.samples"),
    )
    if result.samples <= 0:
        raise M8L2AnalysisConfigError("bootstrap.samples must be positive")
    _frozen(result, M8L2AnalysisBootstrap("paired_moving_block", 2000), "bootstrap contract")
    return result


def _parse_signed_impact(raw: object) -> M8L2AnalysisSignedImpact:
    table = _mapping(raw, "signed_impact")
    _exact_keys(table, _SIGNED_IMPACT_KEYS, "signed_impact")
    result = M8L2AnalysisSignedImpact(
        metric=_text(table["metric"], "signed_impact.metric"),
        side_rule=_text(table["side_rule"], "signed_impact.side_rule"),
        price_rule=_text(table["price_rule"], "signed_impact.price_rule"),
    )
    _frozen(
        result,
        M8L2AnalysisSignedImpact(
            "ofi_signed_future_mid_markout",
            "sign_of_horizon_matched_ofi",
            "ofi_sign_times_future_log_mid_return_bps",
        ),
        "signed-impact contract",
    )
    return result


def _parse_execution(raw: object) -> M8L2AnalysisExecution:
    table = _mapping(raw, "execution")
    _exact_keys(table, _EXECUTION_KEYS, "execution")
    result = M8L2AnalysisExecution(
        market_orders_only=_boolean(table["market_orders_only"], "execution.market_orders_only"),
        probability_threshold=_number(
            table["probability_threshold"], "execution.probability_threshold"
        ),
        symmetric_probability_thresholds=_boolean(
            table["symmetric_probability_thresholds"],
            "execution.symmetric_probability_thresholds",
        ),
        order_notional_usd=_number(table["order_notional_usd"], "execution.order_notional_usd"),
        max_l1_participation=_number(
            table["max_l1_participation"], "execution.max_l1_participation"
        ),
        inventory_order_multiples=_integer(
            table["inventory_order_multiples"], "execution.inventory_order_multiples"
        ),
        reference_price_fit_role=_role(
            table["reference_price_fit_role"], "execution.reference_price_fit_role"
        ),
        reference_depth_fit_role=_role(
            table["reference_depth_fit_role"], "execution.reference_depth_fit_role"
        ),
        reference_price_statistic=_text(
            table["reference_price_statistic"], "execution.reference_price_statistic"
        ),
        reference_depth_statistic=_text(
            table["reference_depth_statistic"], "execution.reference_depth_statistic"
        ),
        reference_quantity_policy=_text(
            table["reference_quantity_policy"], "execution.reference_quantity_policy"
        ),
        l1_fill_policy=_text(table["l1_fill_policy"], "execution.l1_fill_policy"),
        scenario_reset_policy=_text(
            table["scenario_reset_policy"], "execution.scenario_reset_policy"
        ),
        extra_slippage_bps=_number(table["extra_slippage_bps"], "execution.extra_slippage_bps"),
        liquidate_at_end=_boolean(table["liquidate_at_end"], "execution.liquidate_at_end"),
    )
    if not 0.5 < result.probability_threshold < 1.0:
        raise M8L2AnalysisConfigError("execution.probability_threshold must be between 0.5 and 1")
    if result.order_notional_usd <= 0:
        raise M8L2AnalysisConfigError("execution.order_notional_usd must be positive")
    if not 0.0 < result.max_l1_participation <= 1.0:
        raise M8L2AnalysisConfigError("execution.max_l1_participation must be in (0, 1]")
    if result.inventory_order_multiples <= 0:
        raise M8L2AnalysisConfigError("execution.inventory_order_multiples must be positive")
    if result.extra_slippage_bps < 0:
        raise M8L2AnalysisConfigError("execution.extra_slippage_bps must be nonnegative")
    expected = M8L2AnalysisExecution(
        market_orders_only=True,
        probability_threshold=0.55,
        symmetric_probability_thresholds=True,
        order_notional_usd=100.0,
        max_l1_participation=0.10,
        inventory_order_multiples=10,
        reference_price_fit_role="train",
        reference_depth_fit_role="train",
        reference_price_statistic="train_median_mid_price",
        reference_depth_statistic="train_q05_min_bid_ask_l1_depth",
        reference_quantity_policy=("min_100usd_and_10pct_train_q05_l1_depth_rounded_down_to_lot"),
        l1_fill_policy="fill_up_to_recorded_l1_depth_cancel_remainder",
        scenario_reset_policy="per_symbol_session_endpoint_latency_pair",
        extra_slippage_bps=0.0,
        liquidate_at_end=True,
    )
    _frozen(result, expected, "execution contract")
    return result


def _parse_claims(raw: object) -> M8L2AnalysisClaims:
    table = _mapping(raw, "claims")
    _exact_keys(table, _CLAIM_KEYS, "claims")
    result = M8L2AnalysisClaims(
        allow_capacity_claim=_boolean(table["allow_capacity_claim"], "claims.allow_capacity_claim"),
        allow_realized_execution_claim=_boolean(
            table["allow_realized_execution_claim"], "claims.allow_realized_execution_claim"
        ),
        allow_profitability_claim=_boolean(
            table["allow_profitability_claim"], "claims.allow_profitability_claim"
        ),
    )
    _frozen(result, M8L2AnalysisClaims(False, False, False), "claims contract")
    return result


def _parse_source(path: Path, source: bytes) -> M8L2AnalysisConfig:
    try:
        raw = tomllib.loads(source.decode("utf-8"))
    except (UnicodeDecodeError, tomllib.TOMLDecodeError) as error:
        raise M8L2AnalysisConfigError(f"cannot parse M8 live-L2 analysis TOML: {error}") from error
    root = _mapping(raw, "configuration")
    _exact_keys(root, _TOP_LEVEL_KEYS, "configuration")
    return M8L2AnalysisConfig(
        path=path,
        source_sha256=hashlib.sha256(source).hexdigest(),
        study=_parse_study(root["study"]),
        features=_parse_features(root["features"]),
        endpoints=_parse_endpoints(root["endpoints"]),
        regimes=_parse_regimes(root["regimes"]),
        calibration=_parse_calibration(root["calibration"]),
        bootstrap=_parse_bootstrap(root["bootstrap"]),
        signed_impact=_parse_signed_impact(root["signed_impact"]),
        execution=_parse_execution(root["execution"]),
        claims=_parse_claims(root["claims"]),
    )


def semantic_hash_m8_l2_analysis_config(path: str | Path) -> str:
    """Hash validated semantics; this does not authorize non-frozen source bytes."""

    config_path = Path(path).resolve()
    return _parse_source(config_path, config_path.read_bytes()).semantic_sha256


def load_m8_l2_analysis_config(path: str | Path) -> M8L2AnalysisConfig:
    """Load only the exact reviewed, outcome-blind M8 L2 analysis contract."""

    config_path = Path(path).resolve()
    result = _parse_source(config_path, config_path.read_bytes())
    if result.semantic_sha256 != M8_L2_ANALYSIS_CONFIG_SEMANTIC_SHA256:
        raise M8L2AnalysisConfigError(
            "configuration semantics do not match the code-bound outcome-blind freeze "
            f"{M8_L2_ANALYSIS_CONFIG_SEMANTIC_SHA256}"
        )
    if result.source_sha256 != M8_L2_ANALYSIS_CONFIG_SOURCE_SHA256:
        raise M8L2AnalysisConfigError(
            "configuration bytes do not match the outcome-blind freeze "
            f"{M8_L2_ANALYSIS_CONFIG_SOURCE_SHA256}"
        )
    return result


__all__ = [
    "M8_L2_ANALYSIS_CONFIG_SEMANTIC_SHA256",
    "M8_L2_ANALYSIS_CONFIG_SOURCE_SHA256",
    "M8L2AnalysisBootstrap",
    "M8L2AnalysisCalibration",
    "M8L2AnalysisClaims",
    "M8L2AnalysisConfig",
    "M8L2AnalysisConfigError",
    "M8L2AnalysisEndpoint",
    "M8L2AnalysisExecution",
    "M8L2AnalysisFeatures",
    "M8L2AnalysisRegimes",
    "M8L2AnalysisRole",
    "M8L2AnalysisSignedImpact",
    "M8L2AnalysisStudy",
    "M8L2EndpointDomain",
    "M8L2EndpointUnit",
    "load_m8_l2_analysis_config",
    "semantic_hash_m8_l2_analysis_config",
]