feat(miner-manage): loop-closure summary builder (pre-reentry, read-only) (#4282)

[davion-knight]

Adds buildLoopClosureSummary(sources, options) in packages/gittensory-miner/lib/loop-closure.js — a pure, read-only aggregator that summarizes a completed discover → plan → prepare → manage cycle before the miner loop considers re-entering (idle → discovering again). In the spirit of manage-status.js's collectManageStatus: it never calls GitHub, writes a local store, or decides/performs the re-entry itself.

Design

• Caller-supplied cycle boundary (deliberate, per the issue): options.sinceSeq is the event-ledger seq at the end of the prior cycle, so events with a strictly greater seq are "this cycle" — reusing event-ledger.js's own readEvents({ since }) cursor rather than inventing a new persisted boundary marker.
• Generic type-tally: the event ledger stores an open type vocabulary (phase writers define concrete types), so events are counted generically by type — new phase event types (plans built, PRs prepared/opened, outcomes recorded — landing via sibling issues) surface in the tally automatically without a hardcoded list here.
• Output: per-type event counts + lastSeq (the boundary a caller passes as the next cycle's sinceSeq), queue state by status at cycle end, and the current run-state. Split data-collection from any rendering, mirroring manage-status.js.
• Out of scope (explicit): deciding whether to re-enter, or performing the re-entry (setRunState).

Adds the hand-written loop-closure.d.ts alongside the module, matching the package's per-lib .d.ts convention.

Tests

test/unit/miner-loop-closure.test.ts: invalid-sources rejection, empty cycle, generic multi-type tally + last-seq, the sinceSeq boundary excluding prior-cycle events, unknown fallback for missing event/queue kinds + non-integer seq ignored, run-state present/absent, and determinism. Typecheck clean, 100% branch coverage on the new module.

Closes #4282

[superagent-security]

Superagent didn't find any vulnerabilities or security issues in this PR.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 93.96%. Comparing base (ca5576f) to head (af0ce1c).
⚠️ Report is 31 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #4378   +/-   ##
=======================================
  Coverage   93.96%   93.96%           
=======================================
  Files         401      401           
  Lines       36869    36869           
  Branches    13480    13480           
=======================================
  Hits        34644    34644           
  Misses       1569     1569           
  Partials      656      656           

🚀 New features to boost your workflow:

• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[loopover-orb]

Tip

🟩🟩🟩🟩🟩🟩🟩🟩🟩🟩🟩🟩

✅ Gittensory review result - approve/merge recommended

_{Review updated: 2026-07-09 17:17:36 UTC}

3 files · 1 AI reviewer · no blockers · readiness 100/100 · CI green · clean

✅ Suggested Action - Approve/Merge

• safe to merge

Review summary
This adds a small, pure, read-only aggregator (`buildLoopClosureSummary`) that tallies event-ledger and portfolio-queue state between miner-loop cycles, explicitly out of scope for re-entry decisions — matching the stated design and the `manage-status.js` precedent. The `sinceSeq`/`readEvents({since})` boundary is used correctly against the actual ledger contract (strictly-greater seq, as implemented in `event-ledger.js`), and `lastSeq` correctly seeds from `sinceSeq` so an empty cycle round-trips the boundary unchanged. Tests exercise the real branches (multi-type tally, boundary exclusion, unknown fallback, run-state presence/absence, determinism) rather than fabricating unreachable states, and the PR closes the linked issue #4282 with no schema/migration surface touched.

Nits — 5 non-blocking

• loop-closure.js:24 — the external brief flags ~20 cyclomatic complexity against a threshold of 10 for `buildLoopClosureSummary`; consider splitting the event-tally and queue-tally loops into small named helpers (e.g. `tallyEvents`, `tallyQueue`) to bring this under the repo's complexity bar without changing behavior.
• loop-closure.js:33 — an invalid `options.sinceSeq` (non-integer or negative) silently falls back to `null`, which widens the read to 'everything since the beginning' rather than throwing; since callers are expected to pass back a prior `lastSeq`, a malformed value passed by a future caller would silently produce an inflated cycle summary instead of surfacing the bug — consider throwing `invalid_since_seq` to match the fail-loud style used for `invalid_event_ledger`/`invalid_portfolio_queue`.
• The 'unknown' fallback and non-integer-seq guard in the tally loop (loop-closure.js:39-42) protect against data shapes the real SQLite-backed `event-ledger.js` schema (`seq INTEGER NOT NULL UNIQUE`) can't actually produce — reasonable given the generic `sources` interface, but worth a one-line comment noting it's defensive-for-interface-generality rather than a real observed failure mode, so a future reader doesn't chase a phantom bug.
• Split the event/queue tally loops into named helper functions to reduce `buildLoopClosureSummary`'s branch count and make each tally independently testable (loop-closure.js:24-63).
• Make invalid `sinceSeq` a thrown error rather than a silent null fallback, consistent with how `invalid_event_ledger`/`invalid_portfolio_queue` are already handled (loop-closure.js:33).

Signal	Result	Evidence
Code review	✅ No blockers	1 reviewer
Linked issue	✅ Linked	#4282
Related work	✅ No active overlap found	No same-issue or scoped active PR overlap found.
Change scope	✅ 20/20	Low review scope from cached public metadata (1 linked issue).
Validation posture	✅ 25/25	PR body includes validation/test evidence.
Contributor workload	✅ 10/10	Author activity: 88 registered-repo PR(s), 51 merged, 4 issue(s).
Contributor context	✅ Confirmed Gittensor contributor	davion-knight; Gittensor profile; 88 PR(s), 4 issue(s).
Gate result	✅ Passing	No configured blocker found.

Linked issue satisfaction

Partially addressed
The PR delivers a pure, read-only buildLoopClosureSummary with the caller-supplied sinceSeq boundary, event-ledger tallying, queue snapshot, run-state, and matching unit tests exactly as designed in the issue's boundary discussion — but it omits one of the four explicitly listed inputs: collectManageStatus's managed-PR rows (manage-status.js:59), which the issue calls out by name as a required dat

Review context

• Author: davion-knight
• Role context: outside_contributor
• Public audience mode: oss maintainer
• Lane context: Repository is configured for direct PR review.
• Public profile languages: Rust
• Official Gittensor activity: 88 PR(s), 4 issue(s).
• PR-specific overlap: none found.

Contributor next steps

• Keep the PR focused and include validation evidence before maintainer review.

Signal definitions

• Related work = same linked issue, overlapping active PRs, or title/path similarity.
• Change scope = cached public metadata such as size labels, draft state, and review-burden hints.
• Validation posture = whether the PR provides enough public validation/test evidence for maintainer review.
• Contributor workload = public contributor activity and cleanup pressure, not a repo-wide quality failure.
• Contributor context = public GitHub/Gittensor identity context; non-Gittensor status is not a blocker.

_{🟩 Safe / merged · 🟦 Advisory · 🟨 Held for review · 🟥 Blocked / closed}

---

💰 Earn for open-source contributions like this. Gittensor lets GitHub contributors earn for the work they already do — register to start earning →.

Checked by Gittensory, a quiet PR intelligence layer for OSS maintainers.

• Re-run Gittensory review

[loopover-orb]

Gittensory approves — the gate is satisfied and CI is green.