feat(miner): track real per-issue attempt-history on the portfolio queue

[JSONbored]

Summary

buildAttemptGovernorContext hardcoded convergenceInput to a first-attempt-shaped literal on every call — the Governor's fully-built classifyPortfolioConvergence detector (#4286) has never seen real data.

• portfolio-queue.js gains an additive v3 -> v4 migration: attempts_count, consecutive_failures, reenqueue_count (all INTEGER NOT NULL DEFAULT 0), guarded by the same per-column presence check the leased_at migration already uses
• attempts_count increments on every real claim (dequeueNext/batchClaim)
• consecutive_failures + reenqueue_count increment together on every in_progress -> queued transition that did not reach done (markFailed, reclaimStuckItem) — exactly the "cycling queued → in_progress → queued without ever reaching done" the detector's own header names
• consecutive_failures resets to 0 on markDone; reenqueue_count is a lifetime total, deliberately untouched there
• requeueItem (a manual reopen of already-completed work) touches no counter — a different transition than the stuck-loop cycle
• New getAttemptHistory(repoFullName, identifier, apiBaseUrl) returns a real PortfolioConvergenceInput; an item never enqueued reads the same honest zero-state as a genuine first attempt
• attempt-cli.js reads this item's real history (keyed issue:<n>, matching how portfolio-discovery.js enqueues issue-shaped candidates) and threads it into buildAttemptGovernorContext, which stays pure and just forwards it

Closes #5654

Test plan

• test/unit/miner-portfolio-queue.test.ts — schema migration (fresh v3 file, and a defensive idempotency test for a file that already carries all three columns), counter correctness across claim/reclaim/markFailed/markDone/requeue sequences, the reachedDone invariant, apiBaseUrl scoping, and a REGRESSION proving a genuinely non-convergent item now produces a PortfolioConvergenceInput that classifyPortfolioConvergence classifies as non_convergent
• test/unit/miner-attempt-input-builder.test.ts — real convergenceInput threads through unchanged; omitted falls back to the honest zero-state
• test/unit/miner-attempt-cli.test.ts — the real per-issue read is called with the right key and flows into governor.convergenceInput; the real (non-DI) default is exercised via an isolated tmp store
• test/unit/miner-migrate-cli.test.ts updated for the new schema version (v4)
• npm run test:ci green; npm run test:coverage (unsharded) — 100% branch coverage on every new/changed line, verified against lcov.info
• npm run typecheck, npm run build:miner, npm run test:miner-pack, npm audit --audit-level=moderate all clean

[superagent-security]

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

[cloudflare-workers-and-pages]

Deploying with    Cloudflare Workers

The latest updates on your project. Learn more about integrating Git with Workers.

Status	Name	Latest Commit	Preview URL	Updated (UTC)
✅ Deployment successful! View logs	loopover-ui	3ebe095	Commit Preview URL Branch Preview URL	Jul 13 2026, 10:15 PM

[loopover-orb]

Warning

🟨🟨🟨🟨🟨🟨🟨🟨🟨🟨🟨🟨

⏸️ LoopOver review result - manual review recommended

_{Review updated: 2026-07-13 22:18:26 UTC}

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

⏸️ Suggested Action - Manual Review

• AI review could not be completed: The dual-model AI review did not return a usable verdict for this change.

Review summary
This PR replaces a hardcoded first-attempt-shaped literal in buildAttemptGovernorContext with a real per-issue attempt-history query, adding an additive v3->v4 migration (three counters guarded by the same column-presence check as the existing leased_at migration) and a getAttemptHistory read wired through attempt-cli.js into the pure composer. The counter semantics are well-reasoned and consistently applied: attempts_count increments on both claim paths, consecutive_failures/reenqueue_count increment together on the same in_progress->queued-without-done transition (markFailed and reclaimStuckItem), consecutive_failures resets on markDone while reenqueue_count stays a lifetime total, and requeueItem correctly leaves counters untouched since it's a different (already-done) transition. Tests are substantial and include a genuine regression exercising the real classifyPortfolioConvergence detector end-to-end, plus migration idempotency for both a fresh v3 file and one that already has the new columns.

Nits — 7 non-blocking

• The magic PR/issue reference numbers (Track real per-issue attempt-history on the portfolio queue, feed the Governor's convergenceInput #5654) sprinkled through comments in attempt-input-builder.js and portfolio-queue.js are fine as commit-history breadcrumbs but add noise; not necessary to fix.
• portfolio-queue.js and attempt-cli.js are both approaching/at the ~400-line long-file threshold per the external brief — consider whether portfolio-queue.js's migration list is due for extraction as more migrations accrue.
• attempt-cli.js's getAttemptHistory call hardcodes no apiBaseUrl (defaulting to github.com) per its own comment — worth a follow-up issue once multi-forge support lands there, though correctly documented as a known limitation.
• Consider extracting the migration step functions in portfolio-queue.js into a separate migrations module if more are added, to keep the store file under the long-file threshold.
• The comment density in attempt-input-builder.js/portfolio-queue.js is very high; some of the historical narrative (e.g., referencing the old hardcoded literal) could be trimmed once the change lands, since git history already preserves that context.
• PR author also opened the linked issue — Link an issue that was opened by a different contributor, or provide a rationale for why this self-authored issue represents genuine discovery work.
• AI review could not be completed — The gate is held for a human reviewer rather than passed automatically; it re-evaluates on the next update.

Signal	Result	Evidence
Code review	✅ No blockers	1 reviewer
Linked issue	✅ Linked	#5654
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: 45 registered-repo PR(s), 36 merged, 333 issue(s).
Contributor context	✅ Confirmed Gittensor contributor	JSONbored; Gittensor profile; 45 PR(s), 333 issue(s).
Gate result	⚠️ Not blocking	Advisory; not blocking this PR.
Improvement	✅ Minor	risk: clean · value: minor — Code changes are accompanied by test evidence. LLM value judgment: significant — This closes a real, previously-hardcoded gap in the Governor's convergence detection by wiring genuine per-issue attempt-history data end-to-end from schema through to the classifier, with thorough regression coverage proving the detector now fires correctly on real non-convergent data.

Linked issue satisfaction

Addressed
The PR adds the additive attempts_count/consecutive_failures/reenqueue_count schema migration, wires counter updates through claim/markFailed/reclaimStuckItem/markDone, implements getAttemptHistory returning a real PortfolioConvergenceInput, and threads it through attempt-cli.js into buildAttemptGovernorContext without touching non-convergence.ts's classifier, matching the issue's requirements poi

Review context

• Author: JSONbored
• Role context: owner (maintainer lane)
• Public audience mode: oss maintainer
• Lane context: Repository is configured for direct PR review.
• Public profile languages: Python, TypeScript, Ruby, Go, JavaScript, MDX, Shell, Solidity
• Official Gittensor activity: 45 PR(s), 333 issue(s).
• PR-specific overlap: none found.

Contributor next steps

• Treat this as maintainer-lane context rather than normal contributor-lane activity.

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.

[BETA] Chat with Gittensory

Ask Gittensory a question about this PR directly in a comment — grounded only in the same cached, public-safe facts shown above, never a new claim.

• @gittensory ask <question> answers contribution-quality Q&A with source citations and freshness.
• @gittensory chat <question> answers in natural prose from cached decision-pack facts via local inference (maintainer/collaborator; read-only).
• A plain-language @gittensory mention with a real question is routed to the closest matching read-only command automatically -- no exact syntax required.

Full command reference: https://gittensory.aethereal.dev/docs/gittensory-commands

_{🟩 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 LoopOver, a quiet PR intelligence layer for OSS maintainers.

• Re-run LoopOver review