core/tests/test_engine_identity_lineage.py
Shay 512453b6fc
feat(identity): split engine identity from build provenance (ADR-0220 PR C) (#774)
Removes code_revision from the engine-identity hash: engine_identity is now the
sha256 of the 5 ratified packs ONLY. The build revision is provenance (the
manifest's written_at_revision), not identity — so a behavior-neutral rebuild is
the SAME identity and the always-on daemon no longer flag-day strict-breaks on
every commit (the ADR-0220 defect).

Core:
- core/engine_identity.py: ratified_substrate/compute_engine_identity drop the
  git_revision arg (packs-only); add compute_legacy_engine_identity (reproduces
  the pre-split packs+rev hash, for migration verification), ENGINE_IDENTITY_SCHEME=2,
  IdentityReconciliation enum, and reconcile_loaded_identity — the single source of
  truth for the runtime guard AND the workbench reader.
- chat/runtime.py: the load guard reconciles via scheme. Current scheme -> direct
  packs-only compare. Legacy (code_revision-folded) stamp -> a VERIFYING migration:
  reconstruct the legacy hash from the persisted written_at_revision; a match proves
  packs unchanged (warn + re-stamp, resume, no break) — a mismatch means the packs
  genuinely changed (DIVERGED -> strict-refuse). Preserves 'distinct packs => refuse'.
- engine_state/save_manifest: stamp identity_scheme alongside engine_identity
  (additive-optional; no schema_version bump).
- workbench/readers.py: continuity reader uses the same reconcile (no phantom break
  on legacy checkpoints). cli.py break message reworded (packs, not 'build revision').

Callsite cleanup: drop the now-unused git_revision arg + imports across
always_on.py, evals/l10_always_on/runner.py, workbench/readers.py.

Tests:
- test_identity_provenance_split.py (new): 5 reconcile unit proofs + 3 runtime
  integration proofs incl. the wrong=identity defense (legacy stamp of DIFFERENT
  packs still strict-refuses) and the re-stamp migration.
- test_engine_identity.py: invert the code-revision test (rev no longer changes
  identity); assert the substrate is packs-only.
- Restore 3 lineage tests silently red since ADR-0219 (flat-path _manifest helper
  now resolves the gen-dir).
- Update remaining callsites/monkeypatches for the new signature.

Hygiene: .gitignore now covers the ADR-0219 gen-dir runtime files
(current, gen-*/, session_state.json, proposals.jsonl).

Verified: 59 identity/migration/lineage/workbench + 32 L10 + 76 invariants/cli +
34 smoke pass; serving lane SHAs unchanged (no derivation/reliability_gate touch).
2026-06-15 11:38:04 -07:00

80 lines
3.2 KiB
Python

"""EngineIdentity lineage — L11 Phase 2.
Every checkpoint manifest stamps the content-derived ``engine_identity`` the
engine was running under, plus the ``parent_engine_identity`` of the prior
checkpoint — a git-like lineage chain. With a stable ratified substrate the
identity is constant (identity == parent: one stable life); a ratified change
would make identity != parent (the bump). Across reboot the chain is continuous:
the resumed life's checkpoint descends from the pre-reboot identity.
"""
from __future__ import annotations
from pathlib import Path
from chat.runtime import ChatRuntime
from core.cognition.pipeline import CognitiveTurnPipeline
from core.config import RuntimeConfig
from core.engine_identity import engine_identity_for_config
from engine_state import EngineStateStore
def _manifest(state_dir: Path) -> dict:
# Resolve via the store so the ADR-0219 generation-dir manifest is found
# (the committed manifest lives in gen-NNNN/, not at the flat root).
return EngineStateStore(state_dir).load_manifest() or {}
def _drive(state_dir: Path, n: int, config: RuntimeConfig | None = None) -> ChatRuntime:
runtime = ChatRuntime(config=config or RuntimeConfig(), engine_state_path=state_dir)
pipe = CognitiveTurnPipeline(runtime=runtime)
for _ in range(n):
pipe.run("What causes light?")
return runtime
def test_manifest_stamps_content_derived_engine_identity(tmp_path: Path) -> None:
state_dir = tmp_path / "es"
rt = _drive(state_dir, 1)
m = _manifest(state_dir)
assert m["engine_identity"] == rt._engine_identity
# It IS the content-derived identity of the ratified substrate (packs only).
assert m["engine_identity"] == engine_identity_for_config(RuntimeConfig())
def test_lineage_is_stable_across_a_continuous_run(tmp_path: Path) -> None:
state_dir = tmp_path / "es"
_drive(state_dir, 2) # 2 checkpoints
m = _manifest(state_dir)
# Stable substrate -> identity == parent (the life is one continuous identity).
assert m["engine_identity"] == m["parent_engine_identity"]
def test_lineage_is_continuous_across_reboot(tmp_path: Path) -> None:
state_dir = tmp_path / "es"
rt_a = _drive(state_dir, 2)
id_a = rt_a._engine_identity
# Reboot: a fresh runtime over the same checkpoint dir.
rt_b = ChatRuntime(config=RuntimeConfig(), engine_state_path=state_dir)
# The resumed life inherits the pre-reboot identity as its lineage parent.
assert rt_b._loaded_engine_identity == id_a
assert rt_b._engine_identity == id_a # same substrate -> same identity
pipe = CognitiveTurnPipeline(runtime=rt_b)
pipe.run("What is a concept?")
m = _manifest(state_dir)
assert m["parent_engine_identity"] == id_a # descends from the prior life
assert m["engine_identity"] == id_a
def test_different_identity_pack_is_a_different_identity(tmp_path: Path) -> None:
rt_default = ChatRuntime(
config=RuntimeConfig(), engine_state_path=tmp_path / "a", no_load_state=True
)
rt_precision = ChatRuntime(
config=RuntimeConfig(identity_pack="precision_first_v1"),
engine_state_path=tmp_path / "b",
no_load_state=True,
)
assert rt_default._engine_identity != rt_precision._engine_identity