feat(ingest): #404 Phase 1 — thread parent_decision_id through natural-format ingest

[Knapp-Kevin]

Part of #404 (P1) — Phase 1 of 4 (the issue's self-described "smallest enabler").

What

Natural-format ingest now honors a caller-supplied parent_decision_id. Previously only the internal mapping format honored it; the natural format silently dropped it in _normalize_payload. The adapter already reads mapping["parent_decision_id"] (ledger/adapter.py:1460) and persists it — so the fix is to thread it at the single shared normalization point, which both the daemon write.ingest dispatcher and the MCP fallback run.

Also documents the classifier precedence in _classify_decision_level (item #5): explicit decision_level wins → source-type heuristic → (Phase 2) auto-derivation. No logic change there — precedence already worked (#340).

Files

• contracts.py — IngestDecision.parent_decision_id
• handlers/ingest.py — _normalize_payload threading
• ledger/adapter.py — _classify_decision_level docstring (precedence)
• tests/test_404_natural_format_parent_id.py — sociable test (real adapter over memory://)
• docs/META_LEDGER.md — Entry desync optimization V1 — read-path advisory + V2 implementation guide #56 (governance seal)

Acceptance criteria covered

• ✅ refactor: port interfaces + source_span for drift pipeline #1 — natural-format ingest accepts & persists parent_decision_id
• ✅ decision grounding reuse + coverage loop #5 — explicit decision_level overrides the source-type heuristic

Governance

• Risk L2 — additive contract field; no ledger/schema.py change → no §4.7 codeowner gate.
• Independent architect-reviewer audit: PASS — including the load-bearing daemon-serialization check (raw payload is JSON-serialized pre-transport at handlers/ingest.py:845, so the field rides through and is normalized daemon-side).
• Substantiation: the new test is load-bearing — stashing the threading makes it FAIL (row persists NONE); restoring makes it PASS.
• META_LEDGER Entry desync optimization V1 — read-path advisory + V2 implementation guide #56 seals this cycle (chain verified via qorlogic verify-ledger).

Deferred (later phases of #404)

• Phase 2 — Actor L1-parent synthesis + similarity-attach (threshold 0.85), daemon-side.
• Phase 3 — rejected alternatives as L3 negative-siblings.
• Phase 4 — backfill migration. ⚠️ carries the §4.7 @jinhongkuan schema-codeowner gate, and the issue's "v23→v24" must be renumbered (v24 is already taken by the input_span-dedup migration).

🤖 Authored via Qor-logic SDLC on Claude Code

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on base/target branches other than the default branch.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: 61036959-f347-4715-8689-605f11c443d5

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch feat/404-phase1-natural-format-parent-id

---

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 and usage tips.}

[Knapp-Kevin]

Merge note (ledger sealing). This PR carries META_LEDGER Entry #56. PRs #535 and #536 deferred their ledger entries to avoid forking the Merkle chain (all three branched off the same dev tip, Entry #55).

Cleanest path: merge #534 first (it lands #56); then merge #535/#536 in any order and the deferred entries get reconciled in one follow-up pass (#57/#58, re-chained off the new tip). #524 has no ledger entry (handled pre-seal). Net: merge in whatever order you like — the chain gets fixed up afterward.

[jinhongkuan]

Closing during PR hygiene because this branch is stale/conflicting and this is not a top alpha-readiness item. Hierarchical decision display may still be useful, but it should be re-cut against issue #404 only if it is explicitly in product-line scope for the next release. Prefer a small current-dev PR with clear user-facing value over carrying this stale Phase 1 branch.