Verification — issue-332

The testing plan executed (testing-plan.md, rows T1, T3, T9, T11, T13). Commands run from the repository root at the head of claude/github-issue-332-z3xktv. Every login, ref and path in the tests is a fixture (octocat, nobody, octo/repo, a pytest tmp path). Nothing here needed redaction.

Red → green

The new tests were written first. With cli/the_loop stashed back to 161ba52 (13.7.1) and only the tests in place, the fourteen new tests in the selection fail and nine pre-existing or old-code-compatible ones pass (the absent_since unit test is outside this selection; it fails the same way, AttributeError, and is in the T1 selection below):

text

git stash push -- cli/the_loop
uv run --project cli python -m pytest -q cli/tests/test_poller.py cli/tests/test_poller_integration.py \
  -k "ledger or poll_only or closure_check"
FAILED cli/tests/test_poller.py::test_poll_state_note_closure_check_writes_through
FAILED cli/tests/test_poller.py::test_a_closed_ledger_only_item_is_forgotten_and_never_asked_again
FAILED cli/tests/test_poller.py::test_a_poll_only_record_seen_this_window_is_not_asked
FAILED cli/tests/test_poller.py::test_a_poll_only_record_checked_this_window_is_not_asked
FAILED cli/tests/test_poller.py::test_a_poll_only_record_with_an_unparsable_timestamp_is_due
FAILED cli/tests/test_poller.py::test_a_listed_poll_only_record_is_not_asked
FAILED cli/tests/test_poller.py::test_a_stamped_poll_only_record_is_not_asked
FAILED cli/tests/test_poller.py::test_a_poll_only_record_beside_a_session_record_is_tracked_not_ledger_only
FAILED cli/tests/test_poller.py::test_a_still_open_ledger_only_item_is_dated_not_closed
FAILED cli/tests/test_poller.py::test_an_unanswerable_ledger_only_item_is_dated_not_retried_next_cycle
FAILED cli/tests/test_poller.py::test_ledger_only_records_are_asked_longest_absent_first_up_to_the_cap
FAILED cli/tests/test_poller.py::test_unowned_and_degraded_ledger_only_records_do_not_spend_the_cap
FAILED cli/tests/test_poller.py::test_the_cycle_counts_ledger_checks - KeyErr...
FAILED cli/tests/test_poller_integration.py::test_a_closed_ledger_only_item_is_stamped_after_the_window
14 failed, 9 passed, 227 deselected in 1.49s
git stash pop

Three of the nine pass on the old code by design: test_a_poll_only_record_with_a_future_timestamp_is_not_asked (A2 — the old code asked nothing, and neither does the new) and test_a_record_with_poll_beside_another_section_is_asked_without_a_window ×3 (R2 — the tracked set's rule is unchanged, and pinned).

Task	Red (before the change)	Green
1 the schedule	no `absent_since` / `note_closure_check`; no `ledger_checks` on the summary or on `poll.cycle`	`test_poll_state_absent_since_is_the_later_timestamp`, `test_poll_state_note_closure_check_writes_through`, `test_the_cycle_counts_ledger_checks`; `test_eventlog.py::test_every_emitted_event_type_is_documented`
2 the set and the question	a `poll`-only record was never asked; nothing dated a non-closure	the eleven remaining `test_poller.py` tests above; the integration scenario
3 docs	—	`make lint` (markdownlint over every `*.md`), `test_docs_parity.py`

Existing assertions changed: one — test_a_poll_only_record_is_not_reconciled pinned issue-329's R3.2, the rule this work item relaxes, and is now the pair test_a_poll_only_record_seen_this_window_is_not_asked / test_a_closed_ledger_only_item_is_forgotten_and_never_asked_again. One pre-existing test was made robust, not changed in what it proves: test_one_repository_with_issues_disabled_does_not_blind_the_others (issue-315) asserted the session record right after waiting for the tmux spawn, while the dispatcher registers after it spawns, on its own thread — it failed 3 of 20 runs on the unchanged source, and 0 of 20 once the assertion waits for the record as it waited for the spawn.

Rows T1, T3, T9, T11

text

== T1
uv run --project cli python -m pytest -q cli/tests/test_poller.py -k "ledger or poll_only or closure_check or absent_since"
21 passed, 199 deselected in 0.37s
== T3
uv run --project cli python -m pytest -q cli/tests/test_poller_integration.py -k ledger
3 passed, 27 deselected in 0.60s
== T9
uv run --project cli python -m pytest -q cli/tests/test_poller.py cli/tests/test_poller_integration.py -k "ledger or poll_only or closure_check or absent_since"
24 passed, 226 deselected in 0.47s
== T11
uv run --project cli python -m pytest -q cli/tests/test_eventlog.py cli/tests/test_docs_parity.py
19 passed in 0.94s
== the two poller suites, whole
uv run --project cli python -m pytest -q cli/tests/test_poller.py cli/tests/test_poller_integration.py
250 passed in 3.07s

Row T13 — `make check`

text

make check
uv run ruff check cli hooks
All checks passed!
npx --yes markdownlint-cli2@0.18.1 "**/*.md"
Linting: 1016 file(s)
Summary: 0 error(s)
uv run ruff format --check cli hooks
281 files already formatted
uv run pyright cli
0 errors, 0 warnings, 0 informations
uv run python scripts/validate_config.py
VALID   .the-loop/harness-config.yaml
VALID   skills/the-loop/templates/harness-config.yaml
VALID   cli/the_loop/harness-config.default.yaml
VALID   .the-loop/collaborators.yaml
VALID   skills/the-loop/templates/collaborators.yaml
VALID   .the-loop/cli-config.yaml
VALID   skills/the-loop/templates/cli-config.yaml
uv run --project cli python -m pytest -q cli
3200 passed, 1 skipped in 175.77s (0:02:55)

The dashboard's three commands (cd ui && bun run lint && bun run test && bun run build) were not run: no file under ui/ changed (T2 is n/a).

Security

See security-review.md — abuse cases A1–A4, each closed by a named test; no human sign-off at tier 3.