core/chat/deduction_serve_license.py
Claude 71bf04fb44
feat(provenance,teaching): close the lateral gaps the assessment actually found
Squashes the arc's work into one commit; the workflow-file edit it originally
carried is excluded (see the end of this message).

## Lane 1 — Workbench recorded a proved answer as ungrounded

With deduction_serving_enabled ratified ON (ADR-0256), workbench/api.py's live
chat route builds a bare ChatRuntime(), so the deduction composer decides
Workbench turns and stamps grounding_source="deduction" — but
_coerce_grounding_source carried a hand-copied whitelist of the six pre-arc
labels and silently rewrote anything else to "none". The runtime comment
reasoned this was inert because "REPL turns do not flow through Workbench's
CognitivePipelineRecord path". True, and irrelevant: the traffic flows the
other way. Stale since 2026-07-24.

Scope is one field. workbench/api.py:818 prefers TurnEvent.epistemic_state,
which read epistemic_state_needed — honest. So the UNregistered path degraded
honestly while the hand-copied whitelist asserted a falsehood; a second copy of
a closed enum was worse than no copy. Hence registration AND derivation:
GROUNDING_SOURCES exposes the Literal's members, and the coercion reads it.
workbench-ui badges/tokens/snapshot follow; enumCoverage.test.ts forces atomicity.

## Lane 2 — the ratification ceremony

The discovery loop was instrumented but not closed. teaching/ratification.py
turns a reviewed decision into a chain record, a corpus commit, and a receipt.

Its design turns on one observation: _ratified_rows DROPS unadmissible rows
silently — correct when serving, a trap when ratifying, because the file grows,
the commit lands, and the band count does not move. So the ceremony refuses to
call an append a ratification until it has re-read the curriculum through the
real loader and seen the chain arrive; a non-admitted append is rolled back.
Validation is a pre-flight courtesy, admission is the proof.

Arena queue entry and ledger reseal are deliberately NOT performed (bridge rule
1); the receipt names them. Front door: `core proposal-queue ratify`, a sibling
of `review` rather than a flag on it.

## Lane 3 — structural closures

- ADR-0263 gains rule 5: absence policy is DECLARED in CAPABILITY_LEDGERS, not
  passed at the call site. An AST-matched test fails if a serving path passes
  missing_ok again.
- Deductive suite added WHOLE to the pre-push gate: 285 tests in 29s against
  smoke's 216 in 62s, so no coverage trade was needed.
- Smoke/CI parity assertion made bidirectional. It was one-directional, and had
  drifted.
- test_prior_surface_deduction_binding.py pins correction binding on the
  deduction path. The review's diagnosis did NOT reproduce — hash_surface moves
  in lockstep — so it pins what is there. Mutation-checked.
- Domain-keyed ADR index over 312 flat-numbered files, explicitly partial.
- Arc-close brief template, plus this arc's own brief filled in against it.

## Lanes 4 and 5 — two premises falsified by measurement, one of them mine

Math 4.2: baseline reproduced (correct=5 wrong=0 refused=495); all four named
cases traced to one seam with each gap isolated by one-variable probes. Then the
number that changes the recommendation: the gap blocking case 0000 affects 1
case in 500, the 'than' gap blocking 0001 affects 2. ADR-0251's prohibition on
per-case growth now rests on a count. No reader change made.

CGA: versor_condition is 0.22% of a turn, not the "~10x proof latency" I claimed
— that multiplied an isolated microbenchmark by a call count and compared it to
a single verdict's latency. The real cost is geometric_product at 33,986
calls/turn (~73%) via cga_inner in search paths. The obvious closed form is NOT
bit-exact (954/4000 in f32); backend.vault_recall's serial fold IS (3000/3000,
worst-rel 0) and is the correct target. cargo test could not run —
static.crates.io is denied by the sandbox network policy — so the Rust parity
question stays open and the typestate lane is carried forward, not shipped
uncompiled.

## Not landed: three lines owed to .github/workflows/smoke.yml

The CI smoke gate is narrower than the local one —
test_pack_draft_serve_boundary.py (ADR-0253 INV-33) has been local-only, unseen
because the parity pin checked one direction. The edit was authored and rejected
at push for lacking the `workflow` OAuth scope, so it is recorded as a named,
dated PENDING_IN_CI exception rather than dropped: the assertion still fires on
any new divergence, and a second guard fires once the three land.

[Verification]: pre-push gates all green — smoke 236 passed, warmed_session 10
passed, deductive 285 passed. Ratification 14, ADR index 5, CLI suites 10.
Grounding/epistemic sweep 741 passed 1 skipped. workbench-ui 598 passed across
73 files, tsc -b clean. capability index 11 passed, digest unchanged. Math
holdout correct=5 wrong=0 refused=495. Committed chain corpora byte-unchanged
after the tests that write to them.
Environment caveat: the repo pins requires-python ==3.12.13, which uv cannot
fetch for linux-x86_64, so `uv sync --locked` fails. All Python runs used a
scratch venv on 3.12.11 with declared deps — not the locked universe, not the
full ~12k suite. The pin was left untouched. Re-run on a 3.12.13 host before
treating this as merge evidence.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FduW6Krm3PPQv3P5iwBYtx
2026-07-25 04:51:15 +00:00

67 lines
2.7 KiB
Python

"""Serving-side SERVE license for deduction (deduction-serve arc, Phase 3, ADR-0256).
Reads the **ratified, committed** deduction-serve ledger
(``chat/data/deduction_serve_ledger.json``) and exposes, per propositional
shape-band, whether the FULL serving pipeline (reader → projector → ROBDD
engine) has earned ``Action.SERVE`` under the safe default ceilings
(θ_SERVE=0.99). The engine READS this artifact; it never writes it. The
artifact is the sealed-practice output of
``evals.deduction_serve.practice.runner.seal_ledger`` — its ``content_sha256``
is verified on load, so a hand-edited (un-ratified) ledger is rejected rather
than silently trusted.
The load/verify/gate mechanics live in ``core.ratified_ledger`` (ADR-0263) —
this module is the deduction-serve ADAPTER over that bridge: it names the
artifact, keeps the memoization, and preserves its own public API. Ceilings
stay at the safe defaults (invariant #4 — the engine cannot raise its own bar).
"""
from __future__ import annotations
from functools import lru_cache
from core.ratified_ledger import (
RatifiedLedgerError,
ledger_spec,
load_capability_ledger,
serve_license,
)
from core.reliability_gate import Ceilings, ClassTally, LicenseDecision
_LEDGER_CAPABILITY = "deduction_serve"
_LEDGER_PATH = ledger_spec(_LEDGER_CAPABILITY).path
@lru_cache(maxsize=1)
def load_ratified_ledger() -> dict[str, ClassTally]:
"""Load + verify the ratified deduction-serve ledger → per-band ``ClassTally``.
Raises :class:`RatifiedLedgerError` if the file is absent/malformed or its
recomputed ``content_sha256`` does not match the committed one (tamper-evidence:
only the sealed-practice output is trusted, never a hand-edited ledger).
This capability is registered ``missing_ok=False`` in ``CAPABILITY_LEDGERS``
— it ships with its ledger, so absence is a broken deployment, and that
verdict is the manifest's rather than this call's (ADR-0263 rule 5).
"""
return load_capability_ledger(_LEDGER_CAPABILITY)
def deduction_serve_license(
shape_band: str,
*,
ledger: dict[str, ClassTally] | None = None,
ceilings: Ceilings | None = None,
) -> LicenseDecision | None:
"""The ``Action.SERVE`` license for a propositional ``shape_band``, or ``None``.
``None`` means the band is absent from the ratified ledger (no committed
evidence → never licensed; the caller serves a disclosed/hedged surface, the
safe default). Otherwise the deterministic ``license_for`` verdict under the
safe default ceilings.
"""
ledger = ledger if ledger is not None else load_ratified_ledger()
return serve_license(shape_band, ledger, ceilings=ceilings)
__all__ = ["RatifiedLedgerError", "deduction_serve_license", "load_ratified_ledger"]