REPORT — phase5-run-entry-implementer
STEP: phase5-run-entry (Phase 5: the run-spec over an arbitrary repository, the operator start/inspect/stop CLI, and the evidence-policy binding)
REV: a13aacfe2219e3dbd1b6207a28bbe0f980c78b2c   FINAL: dae836f9bee9c03a7708d4364f381658be70caf1
CHANGED: src/orchestrator/run_entry.py — new module: the fail-closed run-spec loader (`load_run_spec`, `validate_run_spec`), the `experiment_target.build_target` freeze (`freeze_target`), the `RunConfiguration` builder (`run_configuration`), the command executor/controller construction (`build_command_executor`, `build_run`), the root-execute drive with acceptance binding or recorded refusal (`start_run`, `_bind_acceptance`), and the read-only `inspect_run` / `stop_run`.
CHANGED: scripts/orchestrate.py — new operator CLI: `start --spec --workdir --store`, `inspect --run-id --store`, `stop --run-id --store`.
CHANGED: docs/run-entry.md — new normative doc: the run-spec field table with a worked example and the start/inspect/stop workflow.
CHANGED: src/orchestrator/schemas.py — the `EvidencePolicy` embedded reference schema (a catalogue `level` plus a positive `version`) and the optional `RunConfiguration.evidence_policy` field that reaches it.
CHANGED: src/orchestrator/glossary.py — `evidence_policy_levels()` parses `## Evidence policy levels`; the catalogue is not redefined locally.
CHANGED: src/orchestrator/controller.py — `Controller.submit_run` records `evidence_policy` beside `policy_version` in the frozen `run_specification` binding; `AdmissionCapability`'s docstring now says data-only construction is refused (F10).
CHANGED: tests/test_run_entry.py — new slow-marked module, one case (or more) per named shape.
VERIFIED: scripts/check-tests --fast — 961 passed, 66 deselected in 69.50s (0:01:09).
VERIFIED: pytest tests/test_run_entry.py — 21 passed in 1.60s.
VERIFIED: pytest tests/test_harness_schemas.py tests/test_boundary_mutations.py tests/test_categorical_completeness.py — 47 passed in 0.20s.
VERIFIED: pytest tests/test_phase3_project_oracle.py — 5 passed in 212.02s (0:03:32).
VERIFIED: scripts/check-static — exit 0 ("All checks passed!" on the gated BLE001/S110 select; the B/SIM advisory report does not gate).
VERIFIED: scripts/check-docs — exit 0 (328 links, 0 errors).
VERIFIED: scripts/check-scope a13aacfe2219e3dbd1b6207a28bbe0f980c78b2c dae836f9bee9c03a7708d4364f381658be70caf1 docs/research/phase5-run-entry-brief.md — `tripped: PILOT-R1 PILOT-R10`; scope: ok (8 path(s)).
NOT VERIFIED: none — every claim below is covered by a run named above; F10's FIXED line reads `no regression test` because a docstring wording defect has no failing assertion to name.
FINDINGS:

F10 (docstring half) | SEVERITY: P2 | ORIGIN: BRIEF | STATE: fixed
  `AdmissionCapability`'s claim that "construction without the module-private mint nonce is refused" is falsified by `copy.copy` and by `__new__`, inviting a reader to claim a third-mint closure that `PILOT-R10` does not buy.
  FIXED: no regression test — this step's later committed code edit replaces the clause in src/orchestrator/controller.py (commit dae836f) with "data-only construction is refused", matching the Capability shape's "a caller holding only data cannot construct one"; a wording fix has no failing assertion at REV to name, and the change makes no claim beyond the data-only threat model.

SHAPES: the seven named shapes are built and each has a slow-marked case: `Run-spec load shape` (`test_run_spec_load_shape`); `Configuration refusal shape` (`test_configuration_refusal_shape`, twelve parameterised malformed specs, plus `test_run_configuration_schema_is_a_second_boundary` for the schema-wrapper case); `Start shape` (`test_start_shape_authorises_acceptance`, `test_start_shape_records_a_refusal_without_authorising`); `Observe shape` (`test_observe_shape_is_read_only`, which re-opens the store file through a second connection); `Stop shape` (`test_stop_shape_settles_live_attempts`, `test_stop_never_rewrites_a_settled_outcome`); `No-harness-import shape` (`test_no_harness_import_shape`, AST-parsing the module and CLI sources, not `sys.modules`); `Evidence-policy shape` (`test_evidence_policy_shape_binds_the_run`). Nothing probed by hand was left unbuilt; the CLI's JSON printing is exercised through the module functions the CLI calls, not through a subprocess.
TRIPPED: PILOT-R1 PILOT-R10 — copied from check-scope's `tripped:` line, verbatim. Both rows are accepted risks whose trigger globs (`src/orchestrator/controller.py`) matched the diff; the brief carries both, no admitted decomposition payload is mutated, and the F10 docstring fix keeps to the data-only threat model.
PROPOSED ROWS: none (the planner maintains docs/residuals.md).
METHOD FRICTION: none against the method's steps. One informational note: `experiment_target.build_target`'s `budgets` keyword is a per-component derivation allocation, not the run's budget dimensions, so `freeze_target` does not forward the run-spec's `budgets` to it (forwarding them raised `DerivationError: budget names no declared component`); the run's budgets reach the run configuration directly. This is not a deviation from the method.
INHERITED: none — the only received item (F10, docstring half) is fixed by this step.
