Gemini bootstrap for #1324

[stranske]

Keepalive: OFF

Source Issue #1324: [Audit][R4] Add 13F amendment (/A) reconciliation as an explicit tested stage

Base: main
Head: gemini/issue-1324

Source: #1324

Automated Status Summary

Scope

13F filings are amended (13F-HR/A), and the field standard is to reconcile them — keep the most recent filing per manager-quarter and supersede 13F-HR with its 13F-HR/A — so diffs and large-delta alerts don't double-count. MDB ingests amendments (adapters/edgar.py: THIRTEEN_F_FORMS = {"13F-HR", "13F-HR/A", "13-F"}) but there is no reconciliation logic (verified: no latest-per-quarter/supersede handling at e5764a5). So raw filings — including amendments — flow into the diff, which both noises the alerts and compounds the duplicate-holdings idempotency bug #1298.

Tasks

• Add a reconciliation step (in the ingest/diff path) that, per (manager_id, period), selects the latest filing and treats 13F-HR/A as superseding the matching 13F-HR; mark superseded filings so downstream diff/alerts use only the authoritative set.
• Ensure the daily diff and large-delta alerts read the reconciled set, not raw filings.

Acceptance Criteria

• Test: ingest a 13F-HR then a 13F-HR/A for the same manager-quarter with changed holdings; assert the diff/alerts reflect the amended holdings once (not both filings double-counted). Current code double-counts.
• Superseded filings remain queryable (provenance preserved) but are excluded from diffs/alerts.
• Deliberate-break demonstration: disabling the supersede step reproduces the double-count in the test; restoring it passes.

Full Issue Text

Why

13F filings are amended (13F-HR/A), and the field standard is to reconcile them — keep the most recent filing per manager-quarter and supersede 13F-HR with its 13F-HR/A — so diffs and large-delta alerts don't double-count. MDB ingests amendments (adapters/edgar.py: THIRTEEN_F_FORMS = {"13F-HR", "13F-HR/A", "13-F"}) but there is no reconciliation logic (verified: no latest-per-quarter/supersede handling at e5764a5). So raw filings — including amendments — flow into the diff, which both noises the alerts and compounds the duplicate-holdings idempotency bug #1298.

Scope

Add an explicit, tested amendment-reconciliation stage that resolves to one authoritative holdings set per (manager, quarter) before diffing/alerting.

Non-Goals

• Do NOT discard amendments — they supersede, not delete; keep provenance.
• Do NOT change the alert thresholds themselves (separate concern).
• Coordinate with [Audit] EDGAR nightly ingest re-runs from 1970 and appends duplicate holdings #1298 (idempotent holdings) — reconciliation assumes stable filing identity.

Tasks

• Add a reconciliation step (in the ingest/diff path) that, per (manager_id, period), selects the latest filing and treats 13F-HR/A as superseding the matching 13F-HR; mark superseded filings so downstream diff/alerts use only the authoritative set.
• Ensure the daily diff and large-delta alerts read the reconciled set, not raw filings.

Acceptance Criteria

• Test: ingest a 13F-HR then a 13F-HR/A for the same manager-quarter with changed holdings; assert the diff/alerts reflect the amended holdings once (not both filings double-counted). Current code double-counts.
• Superseded filings remain queryable (provenance preserved) but are excluded from diffs/alerts.
• Deliberate-break demonstration: disabling the supersede step reproduces the double-count in the test; restoring it passes.

Implementation Notes

Audit baseline e5764a5. Verified: adapters/edgar.py:22 form set includes 13F-HR/A but no reconciliation exists. Refs: WhaleWisdom amendment/split handling. Part of the expansion roadmap (R4); best done after R3 (edgartools) and alongside #1298. dim-5/6 "amendment reconciliation" gap.

Part of epic #1320.

—
PR created automatically to engage Gemini.

[stranske-automation-bot]

Issue #1324: [Audit][R4] Add 13F amendment (/A) reconciliation as an explicit tested stage

Automated Status Summary

Scope

13F filings are amended (13F-HR/A), and the field standard is to reconcile them — keep the most recent filing per manager-quarter and supersede 13F-HR with its 13F-HR/A — so diffs and large-delta alerts don't double-count. MDB ingests amendments (adapters/edgar.py: THIRTEEN_F_FORMS = {"13F-HR", "13F-HR/A", "13-F"}) but there is no reconciliation logic (verified: no latest-per-quarter/supersede handling at e5764a5). So raw filings — including amendments — flow into the diff, which both noises the alerts and compounds the duplicate-holdings idempotency bug #1298.

Tasks

• Add a reconciliation step (in the ingest/diff path) that, per (manager_id, period), selects the latest filing and treats 13F-HR/A as superseding the matching 13F-HR; mark superseded filings so downstream diff/alerts use only the authoritative set.
• Ensure the daily diff and large-delta alerts read the reconciled set, not raw filings.

Acceptance Criteria

• Test: ingest a 13F-HR then a 13F-HR/A for the same manager-quarter with changed holdings; assert the diff/alerts reflect the amended holdings once (not both filings double-counted). Current code double-counts.
• Superseded filings remain queryable (provenance preserved) but are excluded from diffs/alerts.
• Deliberate-break demonstration: disabling the supersede step reproduces the double-count in the test; restoring it passes.

Full Issue Text

Why

13F filings are amended (13F-HR/A), and the field standard is to reconcile them — keep the most recent filing per manager-quarter and supersede 13F-HR with its 13F-HR/A — so diffs and large-delta alerts don't double-count. MDB ingests amendments (adapters/edgar.py: THIRTEEN_F_FORMS = {"13F-HR", "13F-HR/A", "13-F"}) but there is no reconciliation logic (verified: no latest-per-quarter/supersede handling at e5764a5). So raw filings — including amendments — flow into the diff, which both noises the alerts and compounds the duplicate-holdings idempotency bug #1298.

Scope

Add an explicit, tested amendment-reconciliation stage that resolves to one authoritative holdings set per (manager, quarter) before diffing/alerting.

Non-Goals

• Do NOT discard amendments — they supersede, not delete; keep provenance.
• Do NOT change the alert thresholds themselves (separate concern).
• Coordinate with [Audit] EDGAR nightly ingest re-runs from 1970 and appends duplicate holdings #1298 (idempotent holdings) — reconciliation assumes stable filing identity.

Tasks

• Add a reconciliation step (in the ingest/diff path) that, per (manager_id, period), selects the latest filing and treats 13F-HR/A as superseding the matching 13F-HR; mark superseded filings so downstream diff/alerts use only the authoritative set.
• Ensure the daily diff and large-delta alerts read the reconciled set, not raw filings.

Acceptance Criteria

• Test: ingest a 13F-HR then a 13F-HR/A for the same manager-quarter with changed holdings; assert the diff/alerts reflect the amended holdings once (not both filings double-counted). Current code double-counts.
• Superseded filings remain queryable (provenance preserved) but are excluded from diffs/alerts.
• Deliberate-break demonstration: disabling the supersede step reproduces the double-count in the test; restoring it passes.

Implementation Notes

Audit baseline e5764a5. Verified: adapters/edgar.py:22 form set includes 13F-HR/A but no reconciliation exists. Refs: WhaleWisdom amendment/split handling. Part of the expansion roadmap (R4); best done after R3 (edgartools) and alongside #1298. dim-5/6 "amendment reconciliation" gap.

Part of epic #1320.

[stranske-automation-bot]

@gemini start

Please create commits on this branch, run tests, and keep the PR updated.

[coderabbitai]

Warning

Review limit reached

You’ve reached a temporary PR review limit under our Fair Usage Limits Policy.

Your recent review volume is higher than typical usage, so adaptive limits are currently applied.

Next review available in: 43 minutes

Your organization has reached its usage spending cap. Adjust your spending cap in the billing tab.

How can I continue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

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 configuration

Configuration used: Organization UI

Review profile: ASSERTIVE

Plan: Pro

Run ID: 8a76eab3-025c-4d19-a341-76971fcda66e

📥 Commits

Reviewing files that changed from the base of the PR and between b529538 and 715428e.

📒 Files selected for processing (1)

• agents/gemini-1324.md

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch gemini/issue-1324

---

_{Comment @coderabbitai help to get the list of available commands.}

[Copilot]

Pull request overview

This PR adds a minimal “Gemini bootstrap” marker file for issue #1324, likely to initialize or anchor automation/agent workflows associated with that issue.

Changes:

• Add agents/gemini-1324.md containing a bootstrap HTML comment referencing issue #1324.

[stranske]

Closing stale bootstrap after source issue #1324 reached verifier PASS disposition through merged PR #1381. This PR contains only the Gemini bootstrap handoff and no source implementation to merge.