feat(temporal_spine): flip ingest writer default ON, hold lane (#1064)

[robotrocketscience]

What

Flip the ingest-time temporal-spine writer default from OFF to ON
(is_temporal_spine_write_enabled False → True). The retrieval lane
(is_temporal_spine_enabled) is deliberately left default-OFF.

Operator-directed (2026-07-06): the temporal-spine flip (#1064) was split
so the write-side ships now while the read-side waits for the retrieval
cutover.

Why split the flip

Per #1107, the staged lanes — including use_temporal_spine — are wired
into retrieve_v2() only; the production retrieve() hook path does not
expose them until the cutover (garsecg's #1107, Phase 2). So flipping the
lane on now would change bench/eval consumers but not what a live
host retrieves.

The writer, by contrast, is on the live ingest path
(ingest.py → write_temporal_spine), so flipping it default-ON:

• starts chaining per-session TEMPORAL_NEXT edges on every ingest, and
• arms the G4 auto-once backfill (feat(temporal_spine): G4 migration — auto-once backfill + reversible spine clear (#1064) #1090) to build historical spine on the
  first aelf setup after upgrade.

Flipping the writer now lets stores accumulate a dense, production-proven
spine ahead of the read-side flip, de-risking #1107 Phase 2 (temporal
spine is its pilot lane). No retrieval behaviour changes for live hosts
until that cutover.

Gates

Every pre-registered #1064 flip gate is satisfied: G1 (+14.6pp LoCoMo
coverage), G2 (production-budget trim survival + 1986/1986 top-rank
invariance), G3 (latency in-band), G5 (two-build determinism), and G4
migration (#1090, on main).

Blast radius

The writer flip is contained: a store with no TEMPORAL_NEXT edges is
unaffected, and the retrieval lane stays off so no ranking changes. Full
suite 5826 passed / 71 skipped / 75 xfailed. Per-flag opt-out
unchanged (AELFRICE_TEMPORAL_SPINE_WRITE=0 / [ingest] write_temporal_spine = false).

Tests updated to the default-on contract (resolver, ingest path, G4
auto-backfill gate), each with explicit opt-out coverage added; the
non-decisive env / malformed-TOML tests now assert against the
explicit=False lower rung so they still isolate non-decisiveness.

Relationship to #1092

This carves the writer half out of #1092 (the "flip both" HOLD draft).
After this merges, #1092 reduces to the lane-only flip, which folds
into #1107 Phase 2. Coordination note posted on #1092 and #1064.

Refs #1064. Does not close the umbrella (lane flip + #1107 cutover remain).

[sourcery-ai]

Sorry @robotrocketscience, you have reached your weekly rate limit of 500000 diff characters.

Please try again later or upgrade to continue using Sourcery

[coderabbitai]

Warning

Review limit reached

@robotrocketscience, you've reached your PR review limit, so we couldn't start this review.

Next review available in: 21 minutes

Enable usage-based reviews in Billing to review now. Otherwise, wait until the next included review is available.
You're only billed for reviews past your plan's rate limits ($0.25/file).

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: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: 80627b7c-d06e-4f33-9fd4-0d61efb978fc

📥 Commits

Reviewing files that changed from the base of the PR and between 55e2629 and 64f345c.

📒 Files selected for processing (5)

• CHANGELOG/v3.md
• docs/design/feature-temporal-spine.md
• docs/user/CONFIG.md
• src/aelfrice/temporal_spine.py
• tests/test_temporal_spine.py

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch feat/issue-1064-writer-flip

---

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.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

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

[github-actions]

PR-size soft cap

This PR is over the advisory size threshold:

• 140 changed lines (limit: 200)
• 5 changed files (limit: 3)

Bigger PRs collide with more open work, which under the parallel-session workflow tends to produce repeated attn:merge-conflict cycles (see #602). When practical, split into smaller PRs that each touch a focused surface.

This is advisory only — nothing is blocked. If the size is intentional (large refactor, module removal, generated code), apply the size:override label and this comment will be removed on the next push.

[robotrocketscience]

[claim:review:garsecg:2026-07-06T22:46:54Z]

[robotrocketscience]

Review: APPROVE (operator-authorized writer-half flip confirmed).

Reviewed all axes:

• Source (temporal_spine.py): default fallback rung False → True; env > explicit-kwarg > toml > default precedence preserved. Writer only — retrieval lane (is_temporal_spine_enabled) untouched, stays default-OFF.
• Tests: symmetric opt-out coverage — env=0 / toml=false both force off; ingest chains by default and is suppressed on opt-out; auto-backfill runs once by default and defers on opt-out. The two non-decisiveness tests correctly re-anchor to the explicit=False rung so they still isolate "env/toml did not decide" rather than passing by coincidence against the new default.
• Docs: CHANGELOG + CONFIG.md accurately reflect writer default-ON, lane default-OFF (inert on the live retrieve() path until the epic(retrieval): converge production hook onto retrieve_v2 — staged lanes are not on the live path #1107 cutover).
• Mechanics: FF-clean on main, both commits signed, discretion-clean, full CI green (pytest 3.12/3.13 pass, suite 5826 passed per body).

Blast radius contained: retrieval unchanged for live hosts; writer starts chaining TEMPORAL_NEXT and arms the G4 auto-once backfill. Merging via ready-to-merge.

[robotrocketscience]

[release:review:garsecg:2026-07-06T22:49:54Z]

[github-actions]

merge-train: merged 64f345c → main via FF push.