core/tests/test_cli_test_suites.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

266 lines
9.3 KiB
Python

"""CLI tests for curated pytest suite aliases."""
from __future__ import annotations
import tomllib
from pathlib import Path
import pytest
from core import cli_rust
from core import cli
def test_pyproject_installs_core_console_script_under_uv() -> None:
pyproject = Path(__file__).resolve().parents[1] / "pyproject.toml"
data = tomllib.loads(pyproject.read_text(encoding="utf-8"))
assert data["project"]["scripts"]["core"] == "core.cli:main"
assert data["tool"]["uv"]["package"] is True
package_includes = data["tool"]["setuptools"]["packages"]["find"]["include"]
expected_runtime_packages = {
"benchmarks*",
"calibration*",
"core_ingest*",
"demos*",
"engine_state*",
"evals*",
"packs*",
"scripts*",
"teaching*",
}
assert expected_runtime_packages.issubset(package_includes)
def test_core_test_lists_curated_suites(capsys) -> None:
rc = cli.main(["test", "--list-suites"])
captured = capsys.readouterr()
assert rc == 0
assert "fast" in captured.out.splitlines()
assert "smoke" in captured.out.splitlines()
assert "cognition" in captured.out.splitlines()
assert "teaching" in captured.out.splitlines()
assert "packs" in captured.out.splitlines()
assert "deductive" in captured.out.splitlines()
assert "full" in captured.out.splitlines()
def test_cli_smoke_suite_covers_ci_smoke_gate() -> None:
"""Local-first parity pin: the CLI ``smoke`` suite must cover every test
path the CI smoke gate (.github/workflows/smoke.yml) runs.
AGENTS.md makes ``core test --suite smoke`` the mandatory pre-push gate;
if the CI yaml gains a path the CLI tuple lacks, the local gate silently
narrows and regressions clear it — this pin makes that divergence loud.
"""
from core.cli_test import TEST_SUITES
root = Path(__file__).resolve().parents[1]
workflow = (root / ".github/workflows/smoke.yml").read_text(encoding="utf-8")
ci_paths: set[str] = set()
for token in workflow.split():
token = token.strip("\\\"'")
if not token.startswith("tests/"):
continue
if "*" in token:
matches = sorted(root.glob(token))
assert matches, f"CI smoke glob {token!r} matches no files"
ci_paths.update(str(m.relative_to(root)) for m in matches)
else:
ci_paths.add(token)
assert ci_paths, "no tests/ paths parsed from smoke.yml — pin needs updating"
local_paths = set(TEST_SUITES["smoke"])
missing_locally = ci_paths - local_paths
assert not missing_locally, (
"CLI smoke suite is narrower than the CI smoke gate; add to "
f"core/cli_test.py TEST_SUITES['smoke']: {sorted(missing_locally)}"
)
# The other direction, which this pin did not check until 2026-07-25 and
# which had in fact drifted: tests/test_pack_draft_serve_boundary.py sat in
# the local gate and not in smoke.yml. The list's own comment promises the
# two are *equal* ("so the local-first pre-push gate equals the CI gate
# rather than silently narrowing it"), and a one-directional assertion can
# only ever hold half of that. A file only the local gate runs is a file
# whose regression is invisible to anyone reading a green CI badge.
#
# PENDING_IN_CI is a named, dated exception list, not a suppression. These
# three are already in the local gate and belong in smoke.yml; the edit was
# authored and could not be pushed from the authoring session, whose
# credential lacked the `workflow` OAuth scope. Anyone with that scope
# closes this by adding the three lines and emptying this tuple. The
# assertion below still fires on any NEW divergence, which is the
# protection that was missing entirely before.
PENDING_IN_CI = (
"tests/test_pack_draft_serve_boundary.py",
"tests/test_prior_surface_deduction_binding.py",
"tests/test_workbench_deduction_provenance.py",
)
missing_in_ci = local_paths - ci_paths - set(PENDING_IN_CI)
assert not missing_in_ci, (
"CI smoke gate is narrower than the CLI smoke suite; add to "
f".github/workflows/smoke.yml: {sorted(missing_in_ci)}"
)
stale_exceptions = set(PENDING_IN_CI) & ci_paths
assert not stale_exceptions, (
"these are in smoke.yml now — drop them from PENDING_IN_CI: "
f"{sorted(stale_exceptions)}"
)
def test_core_test_suite_expands_to_expected_pytest_paths(monkeypatch) -> None:
calls: list[tuple[str, ...]] = []
def fake_run(*args: str, check: bool = False, cwd=None) -> int:
calls.append(args)
return 0
monkeypatch.setattr(cli, "_run", fake_run)
rc = cli.main(["test", "--suite", "cognition", "-q"])
assert rc == 0
assert calls
command = calls[0]
assert command[:3] == (cli.sys.executable, "-m", "pytest")
assert "tests/test_intent_proposition_graph.py" in command
assert "tests/test_cognitive_turn_pipeline.py" in command
assert "tests/test_articulation_realizer_v2.py" in command
assert "-q" in command
def test_core_test_fast_suite_expands_to_iteration_lane(monkeypatch) -> None:
calls: list[tuple[str, ...]] = []
def fake_run(*args: str, check: bool = False, cwd=None) -> int:
calls.append(args)
return 0
monkeypatch.setattr(cli, "_run", fake_run)
rc = cli.main(["test", "--suite", "fast", "-q"])
assert rc == 0
command = calls[0]
assert "tests/test_cli_test_suites.py" in command
assert "tests/test_runtime_config.py" in command
assert "tests/test_core_semantic_seed_pack.py" in command
assert "tests/test_cognitive_eval_harness.py" in command
assert "tests/" not in command
def test_core_test_deductive_suite_expands_to_entailment_lane(monkeypatch) -> None:
"""Contains-style, not exact-tuple: the ``deductive`` suite has grown
across the deduction-serve/curriculum-serve arcs (v1 through v6-EX,
curriculum, the ratified-ledger bridge, the vocab-trigger instrument) and
keeps growing with each new band — an exact-tuple pin here would go stale
every time, which is exactly what happened before this fix (a pinned
single-file tuple, silently wrong once the suite reached 12+ files, red
outside every real gate — the same leaky-gate class the smoke-suite
coverage pin above exists to prevent)."""
calls: list[tuple[str, ...]] = []
def fake_run(*args: str, check: bool = False, cwd=None) -> int:
calls.append(args)
return 0
monkeypatch.setattr(cli, "_run", fake_run)
rc = cli.main(["test", "--suite", "deductive", "-q"])
assert rc == 0
command = calls[0]
assert command[:3] == (cli.sys.executable, "-m", "pytest")
assert "tests/test_deductive_logic_entail.py" in command
assert "tests/test_deduction_serve_lane.py" in command
assert "tests/test_curriculum_serve.py" in command
assert "-q" in command
def test_core_test_suite_accepts_pytest_flags_without_separator(monkeypatch) -> None:
calls: list[tuple[str, ...]] = []
def fake_run(*args: str, check: bool = False, cwd=None) -> int:
calls.append(args)
return 0
monkeypatch.setattr(cli, "_run", fake_run)
rc = cli.main(["test", "--suite", "packs", "-q"])
assert rc == 0
# Assert against the live suite definition rather than re-hardcoding the
# (growing) packs file list: the contract under test is that a curated
# suite expands to its files followed by the forwarded "-q", with no "--"
# separator needed.
# Assert against the runtime suite map (core.cli_test.TEST_SUITES), which
# cli.cmd_test actually expands. core.cli._TEST_SUITES is a re-export of
# the same object — do not pin a second stale list.
from core.cli_test import TEST_SUITES
assert calls[0] == (
cli.sys.executable,
"-m",
"pytest",
*TEST_SUITES["packs"],
"-q",
)
def test_core_test_passthrough_still_accepts_arbitrary_pytest_args(monkeypatch) -> None:
calls: list[tuple[str, ...]] = []
def fake_run(*args: str, check: bool = False, cwd=None) -> int:
calls.append(args)
return 0
monkeypatch.setattr(cli, "_run", fake_run)
rc = cli.main(["test", "--", "tests/test_core_semantic_seed_pack.py", "-q"])
assert rc == 0
assert calls[0] == (
cli.sys.executable,
"-m",
"pytest",
"tests/test_core_semantic_seed_pack.py",
"-q",
)
def test_core_rust_test_sets_pyo3_forward_compat(monkeypatch) -> None:
calls: list[tuple[tuple[str, ...], object, dict[str, str] | None]] = []
def fake_which(name: str) -> str | None:
return "/usr/bin/cargo" if name == "cargo" else None
def fake_run(
*args: str,
check: bool = False,
cwd=None,
env: dict[str, str] | None = None,
) -> int:
calls.append((args, cwd, env))
return 0
monkeypatch.setattr(cli_rust.shutil, "which", fake_which)
monkeypatch.setattr(cli, "_run", fake_run)
rc = cli.main(["rust", "test"])
assert rc == 0
assert len(calls) == 1
args, cwd, env = calls[0]
assert args == ("cargo", "test", "--release")
assert cwd == cli_rust.CORE_RS_DIR
assert env is not None
assert env["PYO3_USE_ABI3_FORWARD_COMPATIBILITY"] == "1"
def test_non_test_commands_still_reject_unknown_args() -> None:
with pytest.raises(SystemExit):
cli.main(["pack", "list", "-q"])