core/vault/store.py
Shay ef3181aa01 feat(realize): R1 — structural identity & recall (relation-space + span-free idempotency)
R0 keyed a realized fact by its subject's field versor, which is NOT injective:
two facts about one subject embed to byte-identical versors and collide at inf on
metric recall (proven). R1 adds the missing structural key.

- RealizedRecord/metadata carry ordered `relation_arguments` (the relation-space
  key R0's sorted `entity_names` discards) and a span-free `structure_key`.
- `recall_realized(ctx, subject=/predicate=/content_hash=/structure_key=/
  structure_kind=/entity=)` retrieves realized facts by EXACT structural metadata
  (no metric / ANN), via a new read-only `VaultStore.iter_metadata()` accessor.
- Idempotency now dedups on the span-free `structure_key`, so the same proposition
  told from a different source/offset collapses (R0's span-inclusive content_hash
  could not). Guarded by an ambiguous-entity-name refusal — a wrong=0 defense,
  since `Entity.name` is non-unique in the model (only `entity_id` is enforced).
- `content_hash` retained for provenance + replay_hash; `vault_index` pinned to the
  live deque position.

Design adversarially verified (docs/analysis/REALIZE-R1-DETERMINE-scope-2026-06-06.md);
the false "established pattern" private-access comment is removed in favor of the
public accessor. wrong=0 + versor_condition<1e-6 + exact CGA recall preserved;
vault/store.py adds only a read-only accessor (no normalization). Green: 23 realize
+ 110 invariant/vault + 90 smoke; ruff check clean.
2026-06-06 05:52:49 -07:00

484 lines
20 KiB
Python

"""
VaultStore — exact memory via CGA inner product scan.
No HNSW. No approximate nearest neighbor. No index rebuild.
Recall is exact and deterministic over stored versors. When the query is the
same point that was stored, exact self-match is promoted ahead of metric ties
or CGA-sign artifacts.
Exact self-match uses a hash index (versor bytes -> stored indices) instead of
O(N) np.array_equal scans.
"""
from __future__ import annotations
from collections import deque
from typing import TYPE_CHECKING
import numpy as np
from algebra.backend import vault_recall, vault_recall_batch
from algebra.cga import null_project
from core.array_codec import decode_array, encode_array
from core.epistemic_state import EpistemicState
from core.physics.energy import EnergyClass, EnergyProfile
from teaching.epistemic import ADMISSIBLE_AS_EVIDENCE, EpistemicStatus
if TYPE_CHECKING:
from core.physics.learning import VaultPromotionPolicy
# ADR-0006 §"Integration Points":
# "Vault recall re-activates the region to E2 transiently, then lets it
# cool again."
#
# The vault stores crystallized entries (E0 by ADR-0006's "the vault encodes
# the crystallized form"). On recall, each returned entry is stamped with an
# EnergyProfile declaring the transient E2 re-activation. The cool-down is
# the responsibility of downstream field propagation — once the recalled
# region is no longer being injected into the active field state, the
# FieldEnergyOperator's recency decay naturally takes it back down.
#
# raw=0.50 places the profile mid-E2 band (E2 threshold = 0.37, E3 threshold
# = 0.62). The other fields are conservative defaults; consumers that want
# field-specific energy can recompute via FieldEnergyOperator after
# re-injection.
_VAULT_RECALL_RETHAW_ENERGY = EnergyProfile(
raw=0.50,
energy_class=EnergyClass.E2,
convergence_density=0,
activation_count=1,
last_activation_cycle=0,
coherence_residual=0.0,
aspect_weight=0.0,
anchor_adjacent=False,
)
def _versor_key(F: np.ndarray) -> bytes:
return np.asarray(F, dtype=np.float32).tobytes()
# Metadata values are JSON primitives except for one structured value: a
# ``Proposition`` stored under the ``"proposition"`` key (generate/proposition.py).
# It is tagged on encode and reconstructed on decode. The Proposition import is
# lazy (inside the functions) so vault/store.py stays free of a load-time cycle.
_PROPOSITION_TAG = "__core_proposition__"
def _encode_metadata(metadata: dict) -> dict:
from generate.proposition import Proposition
encoded: dict = {}
for key, value in metadata.items():
if isinstance(value, Proposition):
encoded[key] = {_PROPOSITION_TAG: value.to_dict()}
else:
encoded[key] = value
return encoded
def _decode_metadata(metadata: dict) -> dict:
decoded: dict = {}
for key, value in metadata.items():
if isinstance(value, dict) and _PROPOSITION_TAG in value:
from generate.proposition import Proposition
decoded[key] = Proposition.from_dict(value[_PROPOSITION_TAG])
else:
decoded[key] = value
return decoded
def epistemic_state_for_vault_status(entry_status: EpistemicStatus) -> EpistemicState:
"""Map legacy vault review statuses onto the ratified state taxonomy."""
if entry_status is EpistemicStatus.COHERENT:
return EpistemicState.DECODED
if entry_status is EpistemicStatus.FALSIFIED:
return EpistemicState.CONTRADICTED
if entry_status is EpistemicStatus.SPECULATIVE:
return EpistemicState.UNVERIFIED_POSSIBLE
if entry_status is EpistemicStatus.CONTESTED:
return EpistemicState.AMBIGUOUS
return EpistemicState.EPISTEMIC_STATE_NEEDED
def _status_admits(entry_status: EpistemicStatus, min_status: EpistemicStatus) -> bool:
"""Return True iff `entry_status` is admissible at the `min_status` tier.
FALSIFIED entries are never admissible as evidence regardless of the
requested tier — they carry CONTRADICTED semantics and are retained only
for provenance and Stage-3 inversion (ADR-0021 §3). SPECULATIVE entries
are separately excluded at the COHERENT tier (UNVERIFIED-POSSIBLE semantics
— not yet coherent, but distinct from actively falsified). The
exclusion reason for each status is externally inspectable via
``epistemic_state_for_vault_status``: FALSIFIED→CONTRADICTED,
SPECULATIVE→UNVERIFIED_POSSIBLE, CONTESTED→AMBIGUOUS. If the
admissibility set grows in the future (it should not, per ADR-0021), only
this helper changes.
"""
if entry_status is EpistemicStatus.FALSIFIED:
return False # CONTRADICTED — never evidence regardless of requested tier
if min_status is EpistemicStatus.COHERENT:
return entry_status in ADMISSIBLE_AS_EVIDENCE
return entry_status is min_status
def _parse_entry_status(raw_status: object) -> EpistemicStatus:
if isinstance(raw_status, EpistemicStatus):
return raw_status
if isinstance(raw_status, str):
try:
return EpistemicStatus(raw_status)
except ValueError:
return EpistemicStatus.SPECULATIVE
return EpistemicStatus.SPECULATIVE
class VaultStore:
def __init__(
self,
reproject_interval: int = 100,
max_entries: int | None = None,
):
self._versors: deque[np.ndarray] = deque(maxlen=max_entries)
self._metadata: deque[dict] = deque(maxlen=max_entries)
self._store_count: int = 0
self._reproject_interval = reproject_interval
self._max_entries = max_entries
self._exact_index: dict[bytes, list[int]] = {}
# ADR-0054: cached (N, D) f32 matrix view of the deque, rebuilt
# lazily on the first recall after any mutation. Indexing
# optimisation only — scoring arithmetic is unchanged.
self._matrix_cache: np.ndarray | None = None
def store(
self,
F: np.ndarray,
metadata: dict | None = None,
*,
epistemic_status: EpistemicStatus = EpistemicStatus.SPECULATIVE,
) -> int:
"""Store a versor. Returns its index. Auto-reprojects every N stores.
Every stored entry carries an EpistemicStatus stamped into its
metadata under the ``epistemic_status`` key. The default is
SPECULATIVE — the safe choice per ADR-0021 §3: when in doubt,
the entry is not admissible as evidence. Callers that have
actually performed a coherence judgment must declare it
(``epistemic_status=EpistemicStatus.COHERENT``); pack authority
and source provenance alone are not coherence judgments.
"""
arr = np.asarray(F, dtype=np.float32).copy()
stamped: dict = dict(metadata) if metadata else {}
stamped["epistemic_status"] = epistemic_status.value
stamped["epistemic_state"] = epistemic_state_for_vault_status(epistemic_status).value
will_evict = self._max_entries is not None and len(self._versors) >= self._max_entries
self._versors.append(arr)
self._metadata.append(stamped)
if will_evict:
self._rebuild_index()
else:
idx = len(self._versors) - 1
key = _versor_key(arr)
self._exact_index.setdefault(key, []).append(idx)
self._matrix_cache = None
self._store_count += 1
if self._reproject_interval > 0 and self._store_count % self._reproject_interval == 0:
self.reproject()
return len(self._versors) - 1
def recall(
self,
query: np.ndarray,
top_k: int = 5,
*,
min_status: EpistemicStatus | None = None,
) -> list:
"""
Return top_k closest stored versors by CGA inner product.
Each result: {versor, score, metadata, index}.
When ``min_status`` is None (default), no filter is applied —
every stored entry is eligible. This preserves raw session
lookup behavior: the session needs to see its own turns
regardless of epistemic tier.
When ``min_status`` is set, only entries whose stored
``epistemic_status`` is admissible at that tier are returned.
Inference paths that treat vault hits as *evidence* should pass
``min_status=EpistemicStatus.COHERENT`` so SPECULATIVE,
CONTESTED, and FALSIFIED entries do not silently influence
downstream reasoning (ADR-0021 §3).
"""
if not self._versors or top_k <= 0:
return []
query_arr = np.asarray(query, dtype=np.float32)
# Over-fetch when filtering so the post-filter result still
# has a chance at top_k entries. 4x is a generous heuristic;
# vault sizes are bounded by max_entries anyway.
scan_k = max(top_k * 4, top_k) if min_status is not None else max(top_k, 1)
matrix = self._get_matrix()
ranked = vault_recall(
list(self._versors), query_arr, scan_k,
prebuilt_matrix=matrix,
)
key = _versor_key(query_arr)
exact_indices = self._exact_index.get(key, [])
if exact_indices:
exact_matches = [(i, float("inf")) for i in exact_indices]
seen = set(exact_indices)
ranked = exact_matches + [(i, score) for i, score in ranked if i not in seen]
if min_status is not None:
filtered: list[tuple[int, float]] = []
for i, score in ranked:
entry_status = _parse_entry_status(
self._metadata[i].get("epistemic_status", "speculative")
)
if _status_admits(entry_status, min_status):
filtered.append((i, score))
ranked = filtered
return [
{
"versor": self._versors[i],
"score": float(score),
"metadata": self._metadata[i],
"index": i,
"epistemic_state": epistemic_state_for_vault_status(
_parse_entry_status(self._metadata[i].get("epistemic_status", "speculative"))
).value,
# ADR-0006 §"Integration Points": vault recall re-activates the
# region to E2 transiently. The profile here declares the
# re-activation; cool-down is downstream field propagation's
# responsibility once the entry is no longer injected.
"energy_profile": _VAULT_RECALL_RETHAW_ENERGY,
}
for i, score in ranked[:top_k]
]
def recall_batch(
self,
queries: np.ndarray,
top_k: int = 5,
*,
min_status: EpistemicStatus | None = None,
) -> list[list[dict]]:
"""Recall B queries against the stored versors in one sweep.
Returns one ``list[dict]`` per query in the same shape ``recall``
returns. Exact-self-match promotion, ``min_status`` filtering,
and the descending-score / ascending-index tiebreak rule are
applied per query — semantics are identical to looping
``recall(q, top_k=...)`` over each query, but the underlying
scoring scan is a single component-serial sweep over the
cached matrix. ADR-0054.
"""
Q = np.asarray(queries, dtype=np.float32)
if Q.ndim == 1:
Q = Q[None, :]
if not self._versors or top_k <= 0:
return [[] for _ in range(Q.shape[0])]
matrix = self._get_matrix()
assert matrix is not None # non-empty deque ⇒ matrix is built
scan_k = max(top_k * 4, top_k) if min_status is not None else max(top_k, 1)
batch_ranked = vault_recall_batch(matrix, Q, scan_k)
results: list[list[dict]] = []
for b, ranked in enumerate(batch_ranked):
key = _versor_key(Q[b])
exact_indices = self._exact_index.get(key, [])
if exact_indices:
exact_matches = [(i, float("inf")) for i in exact_indices]
seen = set(exact_indices)
ranked = exact_matches + [
(i, score) for i, score in ranked if i not in seen
]
if min_status is not None:
filtered: list[tuple[int, float]] = []
for i, score in ranked:
entry_status = _parse_entry_status(
self._metadata[i].get("epistemic_status", "speculative")
)
if _status_admits(entry_status, min_status):
filtered.append((i, score))
ranked = filtered
results.append([
{
"versor": self._versors[i],
"score": float(score),
"metadata": self._metadata[i],
"index": i,
"epistemic_state": epistemic_state_for_vault_status(
_parse_entry_status(self._metadata[i].get("epistemic_status", "speculative"))
).value,
# ADR-0006: see recall() for re-thaw semantics.
"energy_profile": _VAULT_RECALL_RETHAW_ENERGY,
}
for i, score in ranked[:top_k]
])
return results
def promote_eligible_entries(self, policy: "VaultPromotionPolicy") -> int:
"""Scan SPECULATIVE entries; promote to COHERENT where policy decides.
For each SPECULATIVE entry that carries stored energy metadata, reconstructs
an EnergyProfile and calls policy.decide(). Entries that pass are updated
to COHERENT in-place (metadata only — versors are unchanged, so
_matrix_cache is not invalidated).
Returns the count of promotions made in this call.
ADR-0148.
"""
from core.physics.energy import EnergyClass as _EnergyClass, EnergyProfile as _EnergyProfile
promoted = 0
for meta in self._metadata:
raw_status = meta.get("epistemic_status", "speculative")
try:
entry_status = (
raw_status
if isinstance(raw_status, EpistemicStatus)
else EpistemicStatus(raw_status)
)
except ValueError:
entry_status = EpistemicStatus.SPECULATIVE
if entry_status is not EpistemicStatus.SPECULATIVE:
continue
# Reconstruct EnergyProfile from stored metadata fields.
# If energy metadata is absent, pass None so the policy returns
# "missing_energy_profile" rather than guessing.
energy: _EnergyProfile | None = None
if (
"energy_raw" in meta
and "energy_class" in meta
and "coherence_residual" in meta
):
try:
ec = _EnergyClass(meta["energy_class"])
energy = _EnergyProfile(
raw=float(meta["energy_raw"]),
energy_class=ec,
coherence_residual=float(meta["coherence_residual"]),
)
except (ValueError, TypeError):
energy = None
decision = policy.decide(energy)
if decision.promote:
meta["epistemic_status"] = EpistemicStatus.COHERENT.value
promoted += 1
return promoted
def reproject(self) -> None:
"""
Re-project all stored versors onto the null cone.
Corrects floating-point drift. Run between turns or asynchronously.
"""
reprojected = deque((null_project(v) for v in self._versors), maxlen=self._max_entries)
self._versors = reprojected
self._rebuild_index()
def _rebuild_index(self) -> None:
self._exact_index = {}
for i, v in enumerate(self._versors):
key = _versor_key(v)
self._exact_index.setdefault(key, []).append(i)
self._matrix_cache = None
def _get_matrix(self) -> np.ndarray | None:
"""Return the cached (N, D) f32 stack of stored versors.
Rebuilds the cache on first call after any mutation. Returns
None when the vault is empty so callers can branch without
constructing a 0-row array. ADR-0054.
"""
if not self._versors:
return None
if self._matrix_cache is None:
self._matrix_cache = np.asarray(self._versors, dtype=np.float32)
return self._matrix_cache
def iter_metadata(self):
"""Yield ``(index, metadata)`` for every stored entry, in deque order.
A read-only, non-mutating view over the stored metadata — NOT a
normalization site (``vault/store.py`` is forbidden from repairing field
state; this only reads). ``index`` is the LIVE deque position, the same
integer ``recall``/``recall_batch`` return; it is authoritative in the
unbounded session tier and provenance-only under bounded-vault eviction.
The yielded dicts are the live metadata objects (not copies) — callers
read them; they must not mutate. Lets structured consumers (e.g. realized-
knowledge recall) scan by metadata without reaching into ``_metadata``.
"""
for i, meta in enumerate(self._metadata):
yield i, meta
@property
def reproject_interval(self) -> int:
return self._reproject_interval
@property
def store_count(self) -> int:
return self._store_count
@property
def max_entries(self) -> int | None:
return self._max_entries
def __len__(self) -> int:
return len(self._versors)
def to_dict(self) -> dict:
"""Serialize the vault to a bit-exact, JSON-safe dict (Shape B+ Phase B).
Pure (de)serialization, NOT normalization (``vault/store.py`` is a
CLAUDE.md forbidden normalization site): the persisted versors are the
exact bytes currently in the store — already null-projected at their last
reproject boundary during the live session — encoded losslessly via the
array codec. The derived ``_exact_index`` and the lazy ``_matrix_cache``
are NOT persisted; they are rebuilt deterministically on load. Metadata
is mostly primitives, with one structured value — a ``Proposition`` under
the ``"proposition"`` key (generate/proposition.py) — handled by
``_encode_metadata`` so the snapshot stays JSON-safe.
"""
return {
"versors": [encode_array(v) for v in self._versors],
"metadata": [_encode_metadata(m) for m in self._metadata],
"store_count": int(self._store_count),
"reproject_interval": int(self._reproject_interval),
"max_entries": self._max_entries,
}
@classmethod
def from_dict(cls, payload: dict) -> "VaultStore":
"""Reconstruct a VaultStore from ``to_dict`` output.
The load path performs NO reprojection / normalization / repair: it
restores the exact persisted versors (bit-identical, so exact CGA recall
is preserved) and rebuilds only the derived ``_exact_index``. The lazy
``_matrix_cache`` is left None and rebuilt on the first recall. This is
the bright line — restoring bytes is not a normalization site.
"""
store = cls(
reproject_interval=int(payload["reproject_interval"]),
max_entries=payload["max_entries"],
)
store._versors = deque(
(decode_array(v) for v in payload["versors"]),
maxlen=store._max_entries,
)
store._metadata = deque(
(_decode_metadata(m) for m in payload["metadata"]),
maxlen=store._max_entries,
)
store._store_count = int(payload["store_count"])
store._rebuild_index() # pure: key -> indices over the restored exact bytes
store._matrix_cache = None # derived; lazily rebuilt on first recall
return store