fix(tests): repair storage-cluster test drift from origin-vocab+clock-guard fixes - #3595
Merged
Conversation
…xtures Fresh devtools verify --all triage on master (polylogue-id4n) found 86 unique failures. This batch fixes the clusters that were either genuine bugs with an obvious correct fix or stale test literals/fixtures that needed updating to match intentional production behavior: - polylogue/storage/repair.py: _empty_session_debris_session_ids crashed with `TypeError: tuple indices must be integers or slices, not str` whenever called with a plain sqlite3.Connection (row_factory not sqlite3.Row) -- the exact shape open_readonly_connection() returns, so this also crashes the real `polylogue maintenance repair --target empty_sessions --preview` CLI path, not only tests. Sets row_factory defensively on entry and restores it on exit. Confirmed pre-existing, tracked as polylogue-9rdky; 10 planner tests still fail on a separate, deeper fixture-completeness gap (see bead update). - tests/infra/storage_records.py + test_assertion_candidate_evidence_disclosure.py: SessionBuilder/_record_to_parsed_session and one hand-built ParsedSession set an explicit `title` without `title_source`, so archive_tiers/archive.py's has_real_title gate (title_source must be ORIGIN/HEURISTIC) always degraded them to the "N msgs" structural-label fallback. Every real parser sets both together; mirrors that here. Fixes 4 test_cli_output_schemas.py failures + the evidence-disclosure test. - tests/unit/storage/test_blob_gc.py: 7 tests called time.time() directly in a `_backdate` helper, tripping the clock_guard. blob_gc.py's own age gate reads the real time.time() in production code (not frozen_clock-interceptable), so these genuinely need the real clock -- opted out via `uses_real_clock` rather than introducing frozen_clock. - tests/unit/core/test_enums.py, tests/unit/cli/test_command_aux_runtime.py, tests/unit/devtools/test_verify.py: snapshot-pin literals not updated for the claude-design-session Origin addition (#3422) and three new lab-policy verify steps (raw-payload-hash-purity, position-derived-identity, raw-authority-frontier-executability). Verification: devtools test on all touched files, 204 passed / 10 failed (the pre-existing planner fixture-completeness gap noted above). Ref polylogue-id4n, polylogue-9rdky Co-Authored-By: Claude <noreply@anthropic.com>
… drift Second batch from the polylogue-id4n fresh triage. All test-only fixes are verdict "test-update, not code-fix" -- production behavior changed intentionally in each case and the change's own PR missed a sibling test file exercising the same code path. - polylogue/storage/sqlite/queries/raw_writes.py: the canonical raw_sessions writer's INSERT column list was missing revision_authority_evidence (added to the DDL by #3568 but never wired into the writer). Binds NULL -- the column is only ever populated later by a dedicated maintenance actuator's UPDATE, never at initial-write time. Fixes test_writer_insert_covers_every_live_raw_sessions_column, which exists specifically to catch this class of drift (its own docstring documents the "28-column" contract this restores). - tests/unit/api/test_facade_contracts.py: test_archive_tiers_api_raw_artifacts_read_source_tier only patched archive_tiers/archive.py's datetime via frozen_clock_modules, but parsed_at is actually stamped by archive_tiers/revision_governance.py's _raw_parse_success_state (real datetime.now(UTC), unpatched). Adds that module to the marker. - tests/unit/pipeline/test_branching.py: two independent stale fixtures. (1) #3484 requires structural evidence (matching cwd/git repository_url) for the codex legacy-continuation fallback; the PR updated its two direct test files but missed this one's _codex_continuation_payload fixture, which carried no cwd/git at all. (2) #3495 defaults every session query's SQL-level `root` filter to top-level-only unless the caller overrides it (intentionally applies to the SessionFilter Python API, not just CLI/MCP); a continuation/sidechain session is a child by definition and can never be root, so is_continuation()/is_sidechain() need an explicit is_root(False) override now. (test_neighbor_candidates.py's same-title-candidate failure turned out to share the earlier title_source fix's root cause and is already green.) - tests/unit/storage/test_hermes_artifact_inspection.py: #3576 added a shared magic-byte chokepoint classifying a recognized-but-unclaimed binary payload (e.g. a generic SQLite lookalike) as RECOGNIZED_UNPARSED/ binary_database instead of falling through to an incidental DECODE_FAILED/unknown. The PR updated its four direct test files but missed this sibling file's identical fixture shape. - tests/unit/storage/test_bulk_delete_guarded.py: manually-seeded delegation_facts fixture row used mapping_state='mapped', which has never been a valid DelegationMappingState value (resolved/unresolved/edge_only/ quarantined) -- silently accepted before #3451 wired a real CHECK constraint via literal_check, now correctly rejected. Fixed to 'resolved'. - tests/unit/core/test_paths.py, tests/unit/core/test_query_fields.py: allowlist/coverage tests not updated for two legitimate new additions (api_auth_token_path/claude_code_todos_path path-layout functions; with_unit_windows, an internal with-projection sub-detail parallel to the already-internal with_unit_fields). Verification: devtools test on all touched files, all green. Ref polylogue-id4n
…terals Two more stale-literal fixes from the polylogue-id4n triage, both "test-update, not code-fix" against intentional production changes whose PR missed this sibling test file: - test_codex_event_stream_contract.py: #3567 routes an unrecognized response_item inner type to a distinct codex_unclassified_response_item bucket instead of passing its raw wire token through verbatim (matching the claude_attachment_unclassified precedent). The event is still surfaced, just under the shared bucket name. - test_dispatch_payloads.py: #3447 (polylogue-vf9x) started materializing every Codex `reasoning` response_item as its own THINKING-block message, even one with empty summary/content (block text=None, so the fact that the model reasoned survives). The long-rollout fixture's per-turn message count was never updated to include the +1 reasoning message per turn (120 -> 140 across 20 turns). Verification: devtools test on both files, all green. Ref polylogue-id4n
…lout) test_live_batch_processor_records_durable_attempt's source_path fixture was session_meta-only (no message content), the same fixture-staleness class #3572 already fixed in test_raw_authority_ledger.py and polylogue-h7y0j already tracks for the raw-authority scale-proof generator: PR #3497's require_positive_conversational_evidence() gate correctly refuses a session whose sole message carries no real content, so this fixture always materialized zero sessions (succeeded_file_count 0, not the asserted 1). Adds a real user message record. (test_end_to_end_hidden_root_file_creation_triggers_ingest flickered failed once during iteration on this file -- confirmed via git stash and repeat runs to be pre-existing host-timing flakiness in a real asyncio.sleep-driven filesystem-watcher test, unrelated to this change; not touched.) Verification: devtools test tests/unit/sources/test_live_watcher.py -- 92 passed. Ref polylogue-id4n
…bytes #3539 (polylogue-u19l) retired splicing a synthetic session_meta header into a Codex append-mode capture's payload before hashing/storing it -- the stored raw blob must stay a literal byte-slice of the live file for live-source byte-identity re-verification to work. Recovered identity now flows as a sidecar hint (_AppendPlan.native_id_hint, applied as the parser's fallback_id at replay time) instead. The PR updated test_live_batch_support.py but missed this sibling test, which still asserted the old spliced-header shape. Verification: devtools test tests/unit/sources/test_live_catchup_planning.py -- 20 passed. Ref polylogue-id4n
…tion The test's minimal, hand-rolled sessions/messages/blocks schema (not the real archive DDL) was missing three columns that read_archive_session_envelope now selects: sessions.reported_cost_usd (#3446, cost wiring), messages.stop_reason, and blocks.tool_result_outcome_unknown_reason. Every prefix-sharing composed-read sample hit `sqlite3.OperationalError: no such column`, caught by _sample_prefix_sharing's broad except and surfaced as "1 sampled prefix-sharing composed reads failed" / composed_messages=None / external_counts_citable=False -- masking the actual DB error behind a generic report-level failure reason. This drift had gone undetected because devtools/lineage_validation.py itself hasn't changed since it was added (#2534); testmon-selected runs never had a reason to re-run this file after later columns were added elsewhere, so it only surfaced on this fresh, non-testmon full run. Verification: devtools test tests/unit/devtools/test_lineage_validation.py -- 5 passed. Ref polylogue-id4n
…nt additions - test_daemon_http_security.py: `polylogue ops api token show` (#3488/#3549) is a new, deliberate "print the daemon API bearer token to its own operator" CLI command, same intentional pattern as the pre-existing browser-capture pairing token_show the scanner already allowlists. Adds the new (file, function, sink) tuple. - test_help_snapshots.py: regenerated via --snapshot-update after confirming the diff is purely additive (new --root/--no-root help text from #3495, new `compare` command in "Other commands"). - test_render_quality_reference.py: #3404 (2026-07-30) reworded the Closure Matrix section's guidance text; the test's exact-string assertion still checked the deleted phrasing. Verification: devtools test on all three files, all green. Ref polylogue-id4n
Three fixtures built their synthetic Codex JSONL payload with an escaped double-backslash (source `\\n`), which Python evaluates to the two-character string backslash+n, not an actual newline byte. Every one of these payloads was therefore a single malformed line concatenating two JSON objects, which fails to parse at all -- "no messages, no positive conversational evidence" (PR #3497's gate correctly refusing content with zero parseable structure, not the actual bug). Two of the three tests assert a materialized session count and failed; the third (test_rebuild_index_persists_durable_pass_receipt_alongside_transaction) never checks session counts so the same latent bug didn't surface there, but is fixed for consistency. Verification: devtools test tests/unit/cli/test_archive_maintenance_cli.py -- 58 passed. Ref polylogue-id4n
The ChatGPT parser has never set ParsedSession.title_source, even when it has a genuine provider title (payload["title"]/["name"]). Every other parser that can produce a real title sets title_source alongside it (drive.py, claude/ai_parser.py, claude/code_parser.py) -- ChatGPT was the one omission. This was silent until #3421 (fix(storage): stop title_source=unknown from defeating the structural label) made archive_tiers/archive.py's has_real_title gate require title_source to be ORIGIN/HEURISTIC. Since then, every ChatGPT-origin session -- titled or not -- has silently degraded to the structural "N msgs" fallback label at read time, discarding the genuine title `sessions.title` already carries. Found via a fresh devtools verify --all: tests/unit/cli/test_plain_cli_snapshots.py's "cli-mixed" real-pipeline-seeded fixture (2 chatgpt-export sessions with real payload titles) rendered "7 msgs"/"5 msgs" instead of their actual titles. Fix mirrors drive.py's pattern exactly: title_source=ORIGIN only when the payload carries a real title/name; the bare native-id fallback this parser falls back to when neither is present must NOT count as title evidence -- exactly the "worse than the UUID it replaces" case the whole title_source system exists to catch. Verification: devtools test tests/unit/sources/test_parsers_chatgpt.py tests/unit/sources/test_dispatch_payloads.py -- 143 passed. Ref polylogue-id4n
…anges All drift here traced to real, intentional, already-merged behavior changes -- none papered over without verifying the responsible commit: - token role-split (in=191/out=0 -> in=100/out=91, total conserved): #3511 (fix(cost): role-split estimate fallback tokens) fixed the word-count cost-estimation fallback to attribute an assistant-role message's estimated tokens to output_tokens instead of dumping every role's estimate into input_tokens. Traced by dumping this fixture's raw messages table (all real input_tokens/output_tokens are 0 -- this archive's every "cli-mixed" session cost figure is the heuristic word-count estimate, confirming the split is the #3511 fix, not new drift). - unidentified_artifacts (status output): new counter field, already wired in polylogue/cli/commands/status.py. - total_unit: "top-level sessions" (read --all total): #3495's new root-only-by-default session-query unit label. - index.db/embeddings.db user_version bumps (20/57): real schema evolution across recent migrations, not test-only drift. Verification: devtools test tests/unit/cli/test_plain_cli_snapshots.py tests/unit/cli/test_cli_output_schemas.py -- 56 passed. Ref polylogue-id4n
…land
Completes polylogue-9rdky properly instead of leaving the residual gap as
a note: the row_factory crash fix alone left all 10 planner tests failing
on a second, deeper problem -- their "empty" sessions had no raw_id at all
(raw_id IS NULL), which count_empty_sessions_sync's classifier gate
(_raw_artifact_positively_fails_classification) treats as "no evidence
either way" and therefore never counts as debt. affected_rows read 0
where 1-3 was expected.
Adds DbFactory.mark_as_phantom_debris(native_id) to tests/infra/storage_records.py,
mirroring test_empty_session_repair_provenance.py's existing `_seed` helper
pattern (an agent-*.meta.json-shaped phantom raw artifact, the exact shape
the classifier positively refuses): writes the phantom blob through a
BlobStore scoped to the factory's OWN archive root (self.db_path.parent),
never ambient POLYLOGUE_ARCHIVE_ROOT/blob_store_root() -- test_planner_contract.py's
two callers deliberately seed a "caller" archive root distinct from the
ambient fixture root to prove config-supplied paths win over ambient
defaults, so the phantom blob must land in the same archive the classifier
will actually read back from, not wherever ambient config happens to point.
Also asserts the raw_id UPDATE actually touched exactly one row instead of
silently no-oping on a native_id mismatch (SessionBuilder always stores
native_id as f"ext-{id}", not the bare id callers pass to create_session --
caught this the hard way via a first attempt that updated zero rows).
Verification: devtools test tests/unit/maintenance/ tests/unit/storage/test_empty_session_repair_provenance.py
-- 254 passed. mypy --strict on all three touched files -- clean.
Ref polylogue-id4n, polylogue-9rdky
|
Warning Review limit reached
Next review available in: 48 minutes Enable usage-based reviews in Billing to review now. Otherwise, wait until the next included review is available. How can I continue?After more reviews become available, a review can be triggered using the To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews. How do review limits work?CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability. For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window. Please refer docs for additional details. Review details⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: ASSERTIVE Plan: Pro Plus Run ID: 📒 Files selected for processing (2)
📝 WalkthroughWalkthroughThe changes preserve parsed title provenance, harden SQLite repair state handling, initialize raw-session evidence fields, add phantom-debris test setup, and update Codex, CLI, storage, security, and validation test contracts. ChangesParser, storage, and contract updates
Estimated code review effort: 3 (Moderate) | ~25 minutes Possibly related PRs
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✨ Finishing Touches 💡 1📝 Generate docstrings 💡
🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
Sinity
added a commit
that referenced
this pull request
Aug 3, 2026
Co-Authored-By: Claude <noreply@anthropic.com>
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 2
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@tests/unit/sources/test_dispatch_payloads.py`:
- Around line 720-727: Strengthen the assertions in the per-turn dispatch
payload test around the reasoning response items so it directly verifies that
each turn produces a THINKING-block message, including the reasoning record with
an empty summary and no content. Use the parsed message types or THINKING-block
count rather than relying only on the aggregate total_messages assertion, while
preserving the existing total-count check if useful.
In `@tests/unit/sources/test_live_catchup_planning.py`:
- Around line 297-304: Strengthen the assertions for plan.payload in the live
catch-up planning test to verify it exactly matches the expected append byte
slice from the source fixture, rather than checking for selected substrings.
Keep the native_id_hint assertion and remove the insufficient session_meta/new
substring checks.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: ASSERTIVE
Plan: Pro Plus
Run ID: 419ea145-837c-4cd8-9a45-e09dfdd3272f
📒 Files selected for processing (27)
polylogue/sources/parsers/chatgpt.pypolylogue/storage/repair.pypolylogue/storage/sqlite/queries/raw_writes.pytests/infra/storage_records.pytests/unit/api/test_assertion_candidate_evidence_disclosure.pytests/unit/api/test_facade_contracts.pytests/unit/cli/__snapshots__/test_help_snapshots.ambrtests/unit/cli/__snapshots__/test_plain_cli_snapshots.ambrtests/unit/cli/test_archive_maintenance_cli.pytests/unit/cli/test_command_aux_runtime.pytests/unit/core/test_enums.pytests/unit/core/test_paths.pytests/unit/core/test_query_fields.pytests/unit/daemon/test_daemon_http_security.pytests/unit/devtools/test_lineage_validation.pytests/unit/devtools/test_render_quality_reference.pytests/unit/devtools/test_verify.pytests/unit/maintenance/test_planner_contract.pytests/unit/maintenance/test_planner_filter_narrowing.pytests/unit/pipeline/test_branching.pytests/unit/sources/test_codex_event_stream_contract.pytests/unit/sources/test_dispatch_payloads.pytests/unit/sources/test_live_catchup_planning.pytests/unit/sources/test_live_watcher.pytests/unit/storage/test_blob_gc.pytests/unit/storage/test_bulk_delete_guarded.pytests/unit/storage/test_hermes_artifact_inspection.py
Replace an aggregate message-count-only check with a direct assertion on THINKING-block presence/shape, and replace substring checks on an append payload with an exact byte-slice equality check against the source file's own bytes (the property #3539/polylogue-u19l actually guarantees). Ref polylogue-id4n
Owner
Author
|
Both actionable comments addressed in 2d1335f: THINKING-block presence/shape now asserted directly (not just the aggregate count), and the append-payload check is now an exact byte-slice equality against the source file rather than substring checks. |
Sinity
added a commit
that referenced
this pull request
Aug 3, 2026
Co-Authored-By: Claude <noreply@anthropic.com>
This was referenced Aug 6, 2026
Sinity
added a commit
that referenced
this pull request
Aug 19, 2026
…gbm-quiet, 7qw4) (#4006) ## Summary Retires the four mechanical rows of the WS-B known-red/false-green ledger (`.agent/campaigns/2026-08-overhaul/ws-b-suite-health.md`): uxrim, bwo2l, vqt48, and 7qw4 were all already satisfied on master before this branch was cut; the one row that needed real work was polylogue-ndgbm's quiet component (Codex query-cardinality oracle's fixture never emitted a `message`-shaped record, so the parser's admission gate refused every payload and the module fixture errored at setup), which this PR fixes and whose two associated xfail markers it removes. ## Problem The ledger recorded five findings as known-red or false-green test debt, each with a RECIPE and file:line anchors, dispatched here as a themed sweep. ## Solution - **uxrim** (`tests/unit/api/test_facade_contracts.py:4789`): the recipe asked to extend the `frozen_clock_modules` marker to also cover `revision_governance`. Already done — commit 8ed4f91 (PR #3595, 2026-08-03) landed exactly that extension. No diff needed. - **7qw4** (`tests/unit/storage/test_store_ops.py`): the recipe asked to rewrite the test to call the production `aggregate_message_stats` async fn and delete a test-local shadow reimplementation. Already done — commits from PR #3729 (call the repository path) and PR #3813 (assert on public `origins`) landed this; no shadow function remains in the file. No diff needed. - **bwo2l + vqt48** (`tests/unit/cli/test_archive_maintenance_cli.py`, `tests/unit/sources/test_live_watcher.py`, `tests/unit/sources/test_live_batch_support.py`): the recipe asked to give refused codex fixtures real conversational content. Already done — the 9ykn positive-evidence-gate fixes landed across several prior PRs (visible in `git log --grep`: #3555, #3572, #3600, #3642, #3841). All 138 + 259 tests pass. - **ndgbm quiet component** (`tests/infra/query_manifest_oracle.py`, `tests/unit/cli/test_query_composition_laws.py`): the actual fix in this PR. `QueryCardinalitySession.wire_records()` emitted only `function_call`/`function_call_output` response_items — zero `message`-shaped records — so `codex.is_supported_session_stream()` (`has_message` never set) refused every payload before ingest and the module-scoped fixture errored at setup before either test could run. Added one planted user/assistant message-pair response_item per session (same wire shape landed for the bwo2l/vqt48 fixtures), which admits the stream; action-identity assertions are unaffected because they only count tool rows. Removing the two xfail markers this unblocks surfaced two further pre-existing issues that had never been reachable before (the tests always errored at setup), fixed in the same commit: - The naive-duplicate-id-join mutation test's hand-rolled `actions` view predates the production view's `result_state` CASE projection (polylogue-cuxz.5, `archive_tiers/index.py:1040`) and crashed with `sqlite3.OperationalError: no such column: a.result_state` instead of producing the comparably-wrong row count the test actually asserts on. Added the same CASE projection to the test's naive view. - `find ... then delete --yes --all` has no non-daemon route — it refuses outright when the daemon does not answer the `/prepare` call. Added a daemon-delete stand-in patch on `_submit_daemon_mutation`, mirroring the existing pattern in `test_verb_cardinality.py::TestDeleteCardinalityLargeNonMocked._daemon_delete_route` (does the real delete through `ArchiveStore`, just skips the daemon process). - `_copy_archive`'s plain `shutil.copytree` left the durable bootstrap marker bound to the *original* archive root path (`durable_identity_digest`), so the first write-mode `ArchiveStore.open_existing` at the copied path raised `DurableChangeTrainError: fresh durable bootstrap marker durable identity mismatch`. Re-bind the marker after copy, mirroring `tests/infra/workload_artifacts.py::clone_seeded_archive`. **ndgbm residual (not closed by this PR):** the bead also names a second, separate symptom — the 64-seed process-pool death only reproduces beside a saturated 10-worker gate (confirmed load-only: quiet re-run 20260818T231534Z was green in 43s). That's a load-robustness question the dispatch prompt asked to coordinate with WS-A's e98k cgroup-mmap-budget findings. I could not type this as a `partial` disposition in the machine-readable pr-scope carrier below: the validator requires a **pre-existing Beads dependency edge** (`blocks`/`discovered-from`/`relates-to`/`supersedes`) between ndgbm and its named successor before it will accept the disposition, and no such edge exists between polylogue-ndgbm and polylogue-e98k today. Per the lane contract I do not invoke `bd` from a worktree, so I did not create that link. **Coordinator action needed:** link `polylogue-ndgbm` → `polylogue-e98k` (or a dedicated new successor) via `bd dep add` (or equivalent) before ndgbm can be typed `partial` in a carrier; until then the bead should stay open with its residual noted in its own notes (already present). ## Verification - `devtools test tests/unit/api/test_facade_contracts.py -k raw_artifacts_read_source_tier` → 1 passed (uxrim, unchanged/confirmed) - `devtools test tests/unit/cli/test_archive_maintenance_cli.py` → 138 passed (bwo2l, unchanged/confirmed) - `devtools test tests/unit/sources/test_live_watcher.py tests/unit/sources/test_live_batch_support.py` → 259 passed (vqt48, unchanged/confirmed) - `devtools test tests/unit/storage/test_store_ops.py -k aggregate_message_stats` → 1 passed (7qw4, unchanged/confirmed); mutation red-check locally re-applied AG1 (role-swap on the unfiltered `role_row` query) and AG2 (`words_approx = 0`) against `polylogue/storage/sqlite/queries/stats.py`, both now fail (`assert 1 == 2`, `assert 0 == 9`), reverted before commit — no diff shipped for this row - `devtools test tests/unit/cli/test_query_composition_laws.py` → 2 passed, 0 xfail (ndgbm quiet component: red→green; both `-p no:testmon` and cached runs confirmed) - `devtools verify --quick` → exit 0 (format + lint + mypy + `render all --check`; no `out of sync` in output) All four rows were verified in the context that was failing: each command above is the exact standalone-context command named in the dispatch recipe, run directly (not nested inside a broader suite run). ## Bead disposition matrix | Bead | Disposition | Evidence | Notes | | --- | --- | --- | --- | | polylogue-uxrim | SATISFIED | test pass + commit 8ed4f91 | already fixed on master, PR #3595 | | polylogue-bwo2l | SATISFIED | test pass (138/138) | already fixed on master | | polylogue-vqt48 | SATISFIED | test pass (259/259) | already fixed on master | | polylogue-7qw4 | SATISFIED | test pass + mutation red-check | already fixed on master (PR #3729, #3813) | | polylogue-ndgbm | PARTIAL | test pass (2/2, no xfail) + diff | quiet component fixed this PR; load-only pool-death residual stays open on the bead — **not represented in the typed carrier below** (see Solution section: needs a coordinator-created Beads link to type as `partial`) | <!-- polylogue-pr-scope:v2 { "assigned_beads": [ "polylogue-uxrim", "polylogue-bwo2l", "polylogue-vqt48", "polylogue-7qw4" ], "dispositions": [ { "bead_id": "polylogue-uxrim", "disposition": "satisfied", "evidence": [ { "kind": "test", "ref": "tests/unit/api/test_facade_contracts.py::test_archive_tiers_api_raw_artifacts_read_source_tier" }, { "kind": "commit", "ref": "8ed4f9101813aa142bc915e9c70ef439b86a100d" } ], "successors": [] }, { "bead_id": "polylogue-bwo2l", "disposition": "satisfied", "evidence": [ { "kind": "test", "ref": "tests/unit/cli/test_archive_maintenance_cli.py" } ], "successors": [] }, { "bead_id": "polylogue-vqt48", "disposition": "satisfied", "evidence": [ { "kind": "test", "ref": "tests/unit/sources/test_live_watcher.py" }, { "kind": "test", "ref": "tests/unit/sources/test_live_batch_support.py" } ], "successors": [] }, { "bead_id": "polylogue-7qw4", "disposition": "satisfied", "evidence": [ { "kind": "test", "ref": "tests/unit/storage/test_store_ops.py::test_aggregate_message_stats_reports_public_contract" } ], "successors": [] } ], "mutated_beads": [], "scope_digest": "e397dd57058877a651382c5cf8f681b4cffabc7aae67e285f7192b2c62c787ad", "scope_kind": "bead", "version": 2 } --> Ref polylogue-uxrim, polylogue-bwo2l, polylogue-vqt48, polylogue-ndgbm, polylogue-7qw4 🤖 Generated with [Claude Code](https://claude.com/claude-code) https://claude.ai/code/session_01HWcPJJJvuF25CqVwTFgSQC Co-authored-by: Claude Sonnet 5 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Fresh triage of the storage-cluster test failures tracked by polylogue-id4n: fixes real bugs and stale test drift across raw_sessions column wiring, ChatGPT title-source threading, row_factory handling, and several fixture/snapshot/literal staleness issues from recent origin-vocab and clock-guard additions.
Problem
polylogue-id4n's original count (78 failures) predated PR #3500's full propagation and was known-stale. This branch reran the failing surface fresh, grouped failures by shared root cause, and fixed each cluster: a genuine
raw_sessionscolumn gap, arow_factorycrash, ChatGPT parsertitle_sourcenot being threaded through, several test fixtures asserting stale literals/snapshots against current production behavior, and the maintenance planner's phantom-debris test fixtures not producing real classifiable debris (a deeper bug than the row_factory fix alone addressed — see polylogue-9rdky).Solution
polylogue/storage/repair.py,polylogue/storage/sqlite/queries/raw_writes.py: fix the row_factory crash and wire the missing raw_sessions column.polylogue/sources/parsers/chatgpt.py: threadtitle_sourcethrough the ChatGPT parser.tests/infra/storage_records.py: addDbFactory.mark_as_phantom_debris(native_id), mirroring the existing_seedphantom-artifact pattern, so planner-contract tests produce classifiable debris in the caller's own archive root rather than ambient config.Verification
devtools test tests/unit/maintenance/ tests/unit/storage/test_empty_session_repair_provenance.py— 254 passed.mypy --stricton all touched files — clean.devtools verify --allon this branch — 31 pre-existing failures remain (raw_authority_scale_proof, synthetic_semantics[antigravity], campaign_receipt_reconciliation, etc.), none in files this branch touches. Spot-checked 6 of them directly against origin/master's checkout — identical failures reproduce there. The 4 campaign_receipt_reconciliation failures are a worktree-environment artifact (missing gitignored fixture dir.agent/handoffs/external-agent-campaigns/2026-07-16-gpt-pro-wave, absent from any fresh worktree checkout by design), not a code regression.devtools verify --quick(pre-push gate) — clean.Ref polylogue-id4n
Summary by CodeRabbit
comparecommand.