feat(universe): Phase 4.6 — survivorship-bias fix (historical S&P 500 membership)#274
Merged
Merged
Conversation
… membership)
Closes Research Report v1.0 §7.4 — the highest-priority credibility
fix for top-1% claim. Without this module, every backtest / IC /
PBO / DSR computation using pre-today dates is silently survivorship-
biased: stocks that delisted or got dropped from the S&P 500 are
excluded from the historical view of the universe.
The weekly cron's FORWARD compute is unaffected — current S&P 500
membership remains correct for forward-looking ranking. This fix is
the foundation for HONEST validation in subsequent Phases (5 ML, 6
Lazy Prices, 7 NCO portfolio).
## Changes
- `data/sp500_membership_historical.csv` (NEW, ~55 events) — static
CSV of ADD / REMOVE / RENAME events with per-row Wikipedia or S&P
press-release source URLs. Coverage: 2020-01-01 → present, the
5-year window relevant to current backtest validation. Subsequent
PRs may extend coverage backward.
- `compute/ingest/historical_universe.py` (NEW) —
`members_at(as_of_date, current_universe, anchor_date=None) ->
MembershipResult` pure function. Walks events in reverse from today
back to as_of_date, undoing each ADD/REMOVE that post-dates the
query. Returns `MembershipResult(tickers, as_of_date, anchor_date,
is_complete, events_applied, note)` — `is_complete=False` for dates
before `EARLIEST_EVENT_DATE = 2020-01-01` with a loud warning log
(degraded mode, NOT silent fallback).
- `compute/output/schemas.py` — Rule 18 observability surface:
`Metadata.universe_membership_as_of: str | None` (ISO date of the
historical snapshot used by this compute run; forward cron = today)
+ `Metadata.survivorship_bias_corrected: bool | None` (True when
the lookup was honest, False when degraded to current-only fallback).
- `frontend/lib/types.ts` — TS-side lockstep mirror of the 2 new
fields.
- `frontend/lib/schema-snapshot.json` — regenerated via
`python -m compute.output.schema_check --update-snapshot`.
- `compute/config.py` — `SCHEMA_VERSION` `0.10.6-phase4.5e` →
`0.10.7-phase4.6` (PATCH bump, additive fields, non-breaking).
- `tests/test_config.py` — schema-version pin updated 0.10.6 →
0.10.7-phase4.6 with full rationale docstring.
- `tests/test_ingest/test_historical_universe.py` (NEW) — 20 tests:
- 5 parse-helper edge cases (well-formed / comment / blank /
invalid action / malformed date)
- 2 anchor / future-date guard cases
- 5 known-historical-event regression tests (TSLA 2020-12-21,
SMCI 2024-03-18, ATVI 2023-10-18 acquisition by MSFT, SVB
2023-03-13 collapse)
- 2 pre-coverage degradation cases (is_complete=False below
EARLIEST_EVENT_DATE; exact-boundary case is_complete=True)
- 3 list_known_events accessor tests (sorted, date filter, count
invariant)
- 3 CSV-structure invariants (action in valid set, every row
has http source_url, coverage count >= 30)
## Hard rules preserved
- Rule 9 (schema triple) — atomic Pydantic + TS + snapshot bump ✅
- Rule 18 (observability-before-wiring) — diagnostic surface shipped
in same PR as the module ✅
- Rule 16 (composite formula sacred) — N/A (no scoring change) ✅
- License: factual list (uncopyrightable per Feist 1991 + 17 USC
102(b)); Wikipedia source CC BY-SA 4.0, our compilation = original
work under project MIT license; per-row citation included ✅
- No new pip / npm deps ✅
- No live browser inference ✅
- No FastAPI / Postgres / Docker ✅
- No trade recommendations ✅
## Methodology citations
- Hou, Xue, Zhang (2020). "Replicating Anomalies." Review of
Financial Studies 33(5):2019-2133. Replication crisis evidence
emphasizing survivorship as PRIMARY failure mode in factor-zoo
work.
- McLean, Pontiff (2016). "Does Academic Research Destroy Stock
Return Predictability?" Journal of Finance 71(1):5-32. 32% post-
publication decay; survivorship correction is the first-order
honest-estimate adjustment per the Report's hard-constraint locked
decay budget.
## Verification
- `ruff check compute/ingest/historical_universe.py
tests/test_ingest/test_historical_universe.py` — clean
- `python -m compute.output.schema_check` — Schema snapshot in sync
- `python -m pytest tests/test_config.py::test_schema_version_is_phase4_6
tests/test_ingest/test_historical_universe.py` — 21 passed
- Test suite count: +20 (new offline tests)
## NOT in this PR (deferred to follow-up PRs)
- `compute/validation/pbo_dsr.py` integration to ACCEPT
`universe_provider` kwarg — separate PR after this lands;
unblocks PBO/DSR re-runs with honest historical universes
- `compute/main.py` wiring to populate `Metadata.universe_membership_as_of`
in forward cron — separate PR
- Verify-helper Section M for membership accounting equation
- Extending CSV coverage backward (pre-2020 events)
- Honest re-validation of existing pillars + manipulation_index
using new universe provider (likely revises some published numbers)
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
Contributor
Pre-merge production simulation
Diff vs main
Main baseline: Top-10 movers (sorted by |Δcomposite_score|)
Tickers in main only (2): |
7 tasks
dackclup
added a commit
that referenced
this pull request
May 27, 2026
…bo_dsr gates (#275) Closes the next-task gap surfaced by Research Report v1.0 §7.4 follow-up: PR #274 landed `compute/ingest/historical_universe.members_at()` as a library but nothing called it. This PR threads an optional `universe_provider` callable through `factor_passes_gates()` so the returned metrics dict carries honest universe provenance — closing the loop between the historical-membership module and the validation gates that every Phase 4+ (OSAP / JKP / Qlib / IPCA) and Phase 5 (ML meta-learner) candidate factor must pass. ## What changed - `compute/validation/pbo_dsr.py::factor_passes_gates()` — 3 new optional kwargs (`universe_provider`, `as_of_date`, `current_universe`). When all 3 are passed, the function calls the provider once and enriches the metrics dict with `universe_as_of` (ISO date), `universe_size` (int), and `survivorship_bias_corrected` (bool). When the provider returns `is_complete=False` (e.g., pre-2020 date), the flag flips False AND a warning is logged. When the provider raises, behavior degrades gracefully (3 fields stay None, validation still completes from the supplied returns_matrix). - `compute/validation/pbo_dsr.py::today_utc_date()` — small helper for callers wiring `as_of_date=today_utc_date()` in forward-cron validation paths. - `tests/test_validation/test_pbo_dsr.py` — 6 new tests covering: * backward-compat (no universe kwargs → 3 new keys = None) * happy path (members_at() with date well inside coverage) * degraded path (pre-EARLIEST_EVENT_DATE → is_complete=False) * provider-raises (graceful degradation, validation still runs) * partial-kwargs warning (caller passes only some of the 3) * today_utc_date helper smoke ## Caller migration Pre-Phase-4.6 callers (osap-integration / jkp-integration / qlib scout / ipca scout PRs already merged) are byte-identical — all 3 new kwargs default to None and the function returns the legacy 10-key metrics dict augmented with 3 None values. No caller code needs to change. NEW callers (Phase 4i.1+ integration PRs + Phase 5 ML meta-learner) should pass: from compute.ingest.historical_universe import members_at from compute.ingest.universe import get_sp500_constituents current = frozenset(get_sp500_constituents().ticker) passes, metrics = factor_passes_gates( factor_returns, returns_matrix, n_trials=n_trials, universe_provider=members_at, as_of_date=as_of_date, # backtest cutoff or today_utc_date() current_universe=current, ) The metrics dict can then be threaded into `compute/output/schemas.py::Metadata.universe_membership_as_of` + `Metadata.survivorship_bias_corrected` at writer time (separate follow-up PR — `compute/main.py` wiring stays out of this scope). ## Why this matters (anchor) Hou-Xue-Zhang (2020) RFS replication-crisis evidence emphasizes survivorship as the PRIMARY failure mode in factor-zoo work. Every PBO/DSR number that doesn't carry universe provenance is suspect — pre-Phase-4.6 we couldn't tell, post-Phase-4.6 we can. This PR is the first place that knowledge surfaces in the validation metric dict. ## Hard rules preserved - Rule 9 (schema triple) — N/A; no Pydantic / TS / snapshot touched in this PR. Metadata wiring is deferred to a separate writer PR per scope discipline. - Rule 16 (composite formula sacred) — N/A; no scoring change. - Rule 18 (observability before wiring) — diagnostic surface (metrics dict keys) ships in same PR as the integration; consumer code (writer) lands next, after this metric surface lives in production for ≥ 1 cron cycle. - License: no new deps. Pure forward TYPE_CHECKING import for type hinting `MembershipResult` (zero runtime cost). ## Verification - `ruff check` — clean (linter moved `Callable` to `collections.abc` per modern convention; deliberate, not reverted) - `pytest tests/test_validation/test_pbo_dsr.py tests/test_ingest/test_historical_universe.py tests/test_config.py` — 59 passed in 9.34s - Test count delta: +6 (universe-provider integration cases) - Backward-compat case explicitly tested (no universe kwargs → legacy behavior preserved) ## NOT in this PR (deferred to follow-ups) - `compute/main.py` writer wiring to populate `Metadata.universe_membership_as_of` + `survivorship_bias_corrected` in the forward-cron output JSON - Re-validation of existing pillars + `manipulation_index` with the historical universe (likely revises some published baselines DOWNWARD — explicit honest-correction PR) - Verify-helper Section M for universe-provenance accounting equation Co-authored-by: Claude <noreply@anthropic.com>
4 tasks
dackclup
added a commit
that referenced
this pull request
May 27, 2026
…tadata in forward cron (#276) Closes the last leg of the Phase 4.6 chain. PR #274 landed the `historical_universe.members_at()` module + 2 nullable Metadata fields. PR #275 wired `universe_provider` into `pbo_dsr.factor_passes_gates()` so validation gates carry honest provenance. This PR makes the forward-cron `metadata.json` output ACTUALLY populate those fields instead of leaving them None. ## What changed - `compute/main.py` — `Metadata(...)` construction now passes: - `universe_membership_as_of=now.date().isoformat()` (today's date — forward cron scores as-of today) - `survivorship_bias_corrected=True` (today's S&P 500 IS the honest universe for an as-of-today query, per the PR #274 schema docstring semantic — True means "this output's universe assumption is honest for its as_of_date") - `tests/test_output/test_writer.py` — 2 new round-trip tests: - Phase 4.6 happy path: both fields survive Pydantic → JSON - Legacy snapshot back-compat: when neither field is passed (pre-0.10.7 caller pattern), Pydantic defaults to None and JSON writes nulls ## Hard rules preserved - ✅ Rule 9 (schema triple) — no schema change in this PR (fields already in schemas.py + types.ts + snapshot from PR #274) - ✅ Rule 16 — N/A (no scoring change) - ✅ Rule 18 — observability surface from PR #274 is now actually populated; consumers can branch on it - ✅ No new deps - ✅ No new env-vars ## Verification - `ruff check compute/main.py tests/test_output/test_writer.py` — clean - `python -m compute.output.schema_check` — Schema snapshot in sync - `python -m pytest tests/test_output/test_writer.py -k metadata` — 4 passed (2 existing + 2 new) ## What goes live on next cron Next weekday cron (Wed 2026-05-28 22:00 UTC) writes: metadata.json: ... universe_membership_as_of: "2026-05-28" survivorship_bias_corrected: true Backward compat: legacy snapshots (pre-0.10.7) still have these fields as null per the Pydantic optional default. ## Closes the Phase 4.6 chain | Layer | PR | Status | |---|---|---| | Module | #274 | members_at() + CSV + tests | | Schema | #274 | Metadata fields + types.ts + snapshot | | Validation gate | #275 | universe_provider kwarg in pbo_dsr | | **Writer** | **this PR** | **forward cron populates Metadata** | ## NOT in this PR (next follow-ups) - Honest re-validation of existing pillars + manipulation_index with historical universe (likely shifts PBO/DSR baselines DOWN 5-15%) - Verify-helper Section M for universe-provenance accounting equation - Backtest harness that consumes the new universe_provider end-to-end Co-authored-by: Claude <noreply@anthropic.com>
4 tasks
dackclup
added a commit
that referenced
this pull request
May 27, 2026
…ss (#277) Closes the first leg of honest re-validation per Research Report v1.0 §7.4 task #2. Pure-function + CLI scaffolding that answers the foundational question: **what's the universe drift between today and any historical as-of date?** Future PRs in this chain layer per-pillar IC / PBO / DSR re-baselines on top. ## What ships - `compute/validation/universe_drift.py` (NEW, ~150 LOC) — `compute_universe_drift(as_of_date, current_universe) -> UniverseDriftReport`. Wraps `historical_universe.members_at()` and returns the 3-way symmetric-difference partition: `added_since` / `removed_since` / `unchanged` plus size + completeness diagnostic. `format_drift_report()` renders the human-readable text block with `+N more` cap for long ticker lists. - `scripts/historical_pillar_revalidate.py` (NEW) — CLI wrapper: - `--as-of YYYY-MM-DD` (required) - `--json` for downstream tooling - `--no-fetch-universe` for offline / CI / smoke runs - Exit codes: 0 (clean) / 1 (degraded `is_complete=False`) / 2 (usage) - `tests/test_validation/test_universe_drift.py` (NEW, 11 tests): - Added-since contains recent additions (SMCI, DASH, FSLR, PANW) - Removed-since contains delistings (SVB 2023-03-13) - Anchor-date = zero drift - Pre-EARLIEST_EVENT_DATE = is_complete=False, degraded - Future date raises ValueError - Partition + size invariants (added+unchanged = current; removed+unchanged = historical) - Text rendering contains 4 required section labels - max_listed cap produces "+N more" suffix - Dataclass is frozen - anchor_date default = today UTC - `docs/research/historical-revalidation-harness.md` (NEW) — methodology, CLI usage with sample output, acceptance criteria for next 6 follow-up PRs, caveats ## CLI smoke output (2023-06-01, 7-ticker synthetic universe) ADDED since as_of : 1 tickers SMCI REMOVED since as_of : 9 tickers AAP, ATVI, BIO, BLL, DISH, ETSY, LNC, WHR, ZION ↑ this is the SURVIVORSHIP-BIAS-CORRECTED cohort — current-universe-only views silently EXCLUDE these These 9 are the exact cohort an honest backtest at as-of 2023-06-01 must include. Current-universe-only views (= all pre-Phase-4.6 work) silently dropped them. ## Hard rules preserved - ✅ Rule 9 — no schema change (validation-internal module) - ✅ Rule 16 — N/A (no scoring change) - ✅ Rule 18 — diagnostic surface (UniverseDriftReport dataclass) ships in same PR as the module - ✅ License — no new deps; CSV already on disk from PR #274 - ✅ No frontend touched ## Verification - `pytest tests/test_validation/test_universe_drift.py` — 11/11 pass - `ruff check` — clean - CLI smoke `--no-fetch-universe --as-of 2023-06-01` — exits 0, produces expected 9-ticker REMOVED cohort - CLI degraded `--as-of 2010-01-01` — exits 1, loud warning ## What this PR does NOT do (deferred to next PRs in chain) Per the docs file's "Future-work TODO list": 1. Git-archived `rankings.json` time-series loader (1d) 2. Forward-return computation per ticker from cache (0.5d) 3. Per-pillar IC at historical dates (1d, needs 1+2) 4. PBO/DSR re-baseline via `factor_passes_gates(universe_provider=...)` (1d, needs 3) 5. `manipulation_index` distribution shift report (0.5d, needs 1) 6. `docs/research/honest-baseline-2026-05-27.md` with revised PBO/DSR numbers (0.5d, needs 4) Total to honest-baseline report: ~4-5 days focused dev across a sequence of PRs. ## Methodology citations - Hou, Xue, Zhang (2020). "Replicating Anomalies." Review of Financial Studies 33(5):2019-2133. - McLean, Pontiff (2016). "Does Academic Research Destroy Stock Return Predictability?" Journal of Finance 71(1):5-32. - License: factual list (uncopyrightable per Feist 1991). Co-authored-by: Claude <noreply@anthropic.com>
Merged
5 tasks
dackclup
added a commit
that referenced
this pull request
May 27, 2026
…ignored price cache (#280) New `compute/validation/forward_returns.py` reads the gitignored `compute/cache/prices/<TICKER>.parquet` cache (written by `compute.ingest.prices.fetch_prices`) and computes close-to-close N-month forward total return at any as-of date. Pairs with PR #278's `load_ranking_history()` to close the honest IC re-baseline loop (ranking at T from #2a; realized return at T+horizon from this PR). API: - `compute_forward_return(ticker, as_of_date, horizon_months, *, cache_dir=None) -> float | None` - `compute_forward_return_detailed(...) -> ForwardReturnResult` carrying the actual trading dates / closes / note - `compute_forward_returns_batch(...)` universe batch wrapper - `coverage_report(...)` failure-mode aggregator for the Hou-Xue-Zhang 2020 RFS-style coverage check Edge cases handled: missing parquet → None; no close column → None; as-of doesn't snap within 5d → None; horizon past last cached row → censored = None; start_close ≤ 0 or NaN → None; end_close NaN → None; non-DatetimeIndex parquet → coerced back via `pd.to_datetime`. Source semantics: prefers `Adj Close` (dividend-adjusted) over `Close`; NAIVE returns (no costs / slippage); survivorship-bias correction is NOT done here — callers pair with PR #274's `members_at()` for honest universe construction (Hou-Xue-Zhang 2020 RFS). Tests: 19 new + 1 live-cache smoke (auto-skipped without warm cache). Synthetic OHLCV parquets in `tmp_path` via `cache_dir=tmp_path`. Mirrors PR #278's synthetic-fixture + live-smoke pattern. Schema impact: zero (read-only consumer of existing cache shape). Production-wiring impact: zero (validation tool; no `compute/main.py` hook). #2c per-pillar IC re-baseline will be the first consumer. Honest-baseline disclaimer per Research Report v1.0 autonomous mission constraint: outputs feed IC/DSR/PBO re-baselining, NOT a backtest. Downstream α claims must net frictions (≥30bp/leg), cite McLean- Pontiff 2016 32% decay, and cap at 2-5% net per the ceiling. PHASE_STATUS_INFLIGHT.md updated per PR #237 side-file convention. Harness doc TODO list updated: 4 of 6 items now landed. https://claude.ai/code/session_01AGU8d6pm4u2fQQ5cebg9qa Co-authored-by: Claude <noreply@anthropic.com>
5 tasks
dackclup
added a commit
that referenced
this pull request
May 27, 2026
…tes (#281) New `compute/validation/historical_ic.py` orchestrator pairs PR #278's `load_ranking_history` (ranking at T) with PR #280's `compute_forward_returns_batch` (realized return at T + horizon) and computes per-pillar Spearman IC across the historical window — closes the IC re-baseline half of the Phase 4.6 chain. API: - `compute_pillar_ic(scores, returns, *, method, min_tickers)` pure cross-sectional IC for one (pillar, date) pair - `compute_historical_ic_report(start, end, *, horizon_months, pillars, ...)` walks rankings.json snapshots + forward returns cache, aggregates into `HistoricalICReport` - `format_ic_report(report)` human-readable text rendering - `PillarICEntry` / `PillarICSummary` / `HistoricalICReport` three frozen-dataclass carriers Spearman computed as Pearson on rank-transformed series (Spearman 1904 + Conover 1999 §5.4) to avoid pulling scipy into the dep set (QuantRank ships without scipy; pandas' `Series.corr(method= 'spearman')` requires it transitively). Drops with descriptive notes: - cross-section < MIN_TICKERS_PER_DATE = 30 (Grinold-Kahn 2000 §4.2) - None / NaN / inf in either input - constant inputs (std=0 → correlation undefined) Aggregates per pillar: mean / std / median / min / max / IC IR / hit-rate. IC IR = mean/std × sqrt(n_dates) (Grinold-Kahn 2000 §4.4). Hit-rate = fraction of dates with strictly positive IC. Honest-baseline disclaimer per Research Report v1.0: - IC reported here is NAIVE — no costs / slippage / sector neutralization. Real net-of-cost IC typically 30-50% smaller per McLean-Pontiff 2016 JF post-publication decay - The historical universe MUST come from PR #274 members_at() to avoid survivorship bias (Hou-Xue-Zhang 2020 RFS); orchestrator reads the historical universe FROM rankings.json at as-of T which is correct by construction (snapshot itself is historical universe) - Report is a TIME SERIES + summary, not a single headline number Tests: 28 new (28 passing). Coverage: module constants, pure IC computation edge cases (perfect ±1.0, constant inputs, NaN drops, below-min cross-section, method validation), summary aggregation math (IC IR formula pinned, hit-rate semantics), orchestrator full-path (one date / multi-date / missing pillar / malformed JSON), text rendering, and a live-git smoke that auto-degrades gracefully when the gitignored price cache is absent. Schema impact: zero. No new Pydantic / TS / snapshot field. Production-wiring impact: zero. No compute/main.py import. The orchestrator is purely a validation / re-baseline tool. Downstream PRs (#2d PBO/DSR re-baseline + #2f honest-baseline report) consume the output. Phase 4.6 chain status: 5 of 6 items now landed (#1/#2 PR #277, #2a PR #278, #2b PR #280, #2c this PR, #2e PR #279; #2d gate kwarg shipped PR #275). #4 PBO/DSR re-baseline needs a warm-CI execution to publish actual numbers; #6 honest-baseline doc closes the chain. PHASE_STATUS_INFLIGHT.md updated per PR #237 side-file convention. Harness doc TODO list updated: 5 of 6 items now landed. https://claude.ai/code/session_01AGU8d6pm4u2fQQ5cebg9qa Co-authored-by: Claude <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
2 participants
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.
First executed feature of Research Report v1.0 autonomous pipeline. P0 per Report §7.4 — credibility-bearing prerequisite for every subsequent Phase 5 (ML), 6 (Lazy Prices), 7 (NCO) validation gate.
Why this matters
Without historical S&P 500 membership, every backtest / IC / PBO / DSR computation that uses pre-today dates is silently survivorship-biased: stocks that delisted or got dropped from the index are excluded from the historical view of the universe. Hou-Xue-Zhang (2020) RFS calls this out as the PRIMARY failure mode in factor-zoo replication; McLean-Pontiff (2016) JF locks the 32% post-publication decay budget that ONLY holds under survivorship-corrected estimates.
The weekly cron's forward compute is unaffected — current S&P 500 membership remains correct for forward-looking ranking. This fix is foundational for every honest validation in subsequent Phases.
What landed
data/sp500_membership_historical.csv(NEW, ~55 events)compute/ingest/historical_universe.py(NEW)members_at(as_of_date, current_universe, anchor_date)pure function returningMembershipResult(tickers, is_complete, events_applied, note). Reverses post-as-of events to compute historical universe.compute/output/schemas.pyuniverse_membership_as_of(ISO date) +survivorship_bias_corrected(bool).frontend/lib/types.tsfrontend/lib/schema-snapshot.json--update-snapshot.compute/config.pySCHEMA_VERSION0.10.6-phase4.5e→0.10.7-phase4.6.tests/test_config.pytests/test_ingest/test_historical_universe.py(NEW)Hard rules preserved
Verification
ruff check(new files)python -m compute.output.schema_checkpython -m pytest tests/test_config.py + tests/test_ingest/test_historical_universe.pyKnown-event coverage in test suite
members_at(2020-12-20)excludes TSLA ✅;members_at(2020-12-22)includes TSLA ✅members_at(2024-03-17)excludes SMCI ✅members_at(2023-10-17)includes ATVI ✅ (would be missed by current-universe-only)members_at(2023-03-12)includes SVB ✅ (classic survivorship-bias example)Methodology citations
NOT in this PR (deferred to follow-up PRs)
compute/validation/pbo_dsr.pyintegration to acceptuniverse_providerkwarg — unblocks PBO/DSR re-runs with honest historical universescompute/main.pywiring to populateMetadata.universe_membership_as_ofin forward cronTest plan
ruff check .— clean (new files)python -m compute.output.schema_check— cleanpytest tests/test_ingest/test_historical_universe.py tests/test_config.py— 21/21 passtypes.ts)methodology-scientistsubagent — verify HXZ 2020 + MP 2016 citations (FOLLOW-UP review)schema-sentinelsubagent — verify Pydantic ↔ TS ↔ snapshot triple sync (auto-fired by hook)dependency-auditor— N/A (no new deps)security-reviewer— N/A (no new env-vars, workflows, or surfaces)Subscribe-after-open suggestion: same as PR #271 / #272 / #273 — subscribe me to PR activity for CI + review comments.
Related
.claude/skills/phase-4.6/survivorship-bias-fix/PLAN.mdcommitted in PR docs(research): Calibration + 5 PLAN drafts from Research Report v1.0 #273Generated by Claude Code