refactor(contract): remove CollapseGateEmission + MailboxSoA::emit(); in-place consume_firing() successor#487
Merged
Conversation
#477 three-tier model The PR #477 'what does NOT exist' table, made source-true: - Remove CollapseGateEmission from contract::collapse_gate (+ lib.rs re-export). MailboxId, MergeMode, GateDecision survive — they are addressing/merge concepts, not emission concepts. - Remove MailboxSoA::emit(); rename last_emission_cycle → last_active_cycle. New in-place consume_firing(row) keeps the threshold + same-cycle-idempotency semantics with no carrier object (zero-copy: owner reads row columns in place, then stamps). - Reword 4 stale doc references (kanban, episodic_edges, witness_tombstone, mailbox_soa module header) to the three-tier zero-copy framing. - Supersede the CLAUDE.md Baton-scoping block (dated note; bundle math + mailbox-as-owner safety argument survive untouched). - Fix cycle-coherent-soa-snapshot-v1 D-SOA-SNAP-1/2: generic MailboxSoaSnapshot<C> + SnapshotProvider::Column so contract never imports MultiLaneColumn (closes #477 CodeRabbit Critical; contract stays zero-dep). - Board: LATEST_STATE contract inventory updated, TD-COLLAPSE-GATE- SMALLVEC-1 closed as moot, AGENT_LOG entry prepended. Verified: #477 codex P2 on verify_layout (ColumnOutOfBounds) already fixed on main with regression test column_past_stride_caught. Tests: contract 594 (-8 emission, +2 gate/merge), driver 85 (emit tests rewritten as consume tests, +1 OOB), clippy clean, workspace check clean. https://claude.ai/code/session_017GFLBnDy23AWBqvkbHHC41
|
Warning Review limit reached
More reviews will be available in 41 minutes and 59 seconds. Learn how PR review limits work. Your organization has run out of usage credits. Purchase more credits in the billing tab to continue. ⌛ How to resolve this issue?After more reviews become available, a review can be triggered using the We recommend that you space out your commits to avoid hitting the rate limit. 🚦 How do rate limits work?CodeRabbit enforces hourly rate limits for each developer per organization. Our paid plans include higher PR review limits than trial, open-source, and free plans. In all cases, reviews become available again over time. During sustained high-volume PR review activity, CodeRabbit may temporarily slow when the next review becomes available. Please see our Fair Usage Limits Policy for further information. ℹ️ Review info⚙️ Run configurationConfiguration used: Organization UI Review profile: CHILL Plan: Pro Plus Run ID: 📒 Files selected for processing (2)
📝 WalkthroughWalkthroughThis PR documents and implements the architectural follow-up to PR ChangesThree-Tier Model Implementation & Cleanup
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~25 minutes Possibly related PRs
Poem
🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. 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. Comment |
…ry commit, not a cleanup Reframes PR #487 correctly: removing CollapseGateEmission was not dead-code hygiene, it was a boundary-shape correction. The inner seam (mailbox-to- mailbox) does not exist as a typed crossing — it is ownership transfer under Rust move semantics (E-CE64-MB-4). The outer seam (SoA ↔ Lance) is ontology-mediated: OGAR class = table schema (label + fields + tools + templates) SoaEnvelope = column mapping (byte geometry) Lance I/O = SQL writer (LE bytes from in-place store) hand-DTO = ORM-bypass (CollapseGateEmission was the last) The class supplies semantics; the envelope supplies geometry; Lance does the writing. Any independent carrier struct at this seam is schema drift by definition (per #477's 'every DTO is a derived view of an OGAR class'). MailboxId / MergeMode / GateDecision survive because they are vocabulary *of* the ontology side (addressing, merge policy, gate decision), not parallel descriptions of row data. Consequences for future PRs: - inner seams are moves, never carrier types - the outer seam has exactly one description (class + envelope + Lance) - hand-rolled active records are ORM-bypass — reach for ClassView + FieldMask + class-template specialization instead Cross-refs: PR #477 three-tier model; PR #487 tombstone commit; soa-three-tier-model.md; E-OGAR-NORTHSTAR-1 (class spine); I-LEGACY-API-FEATURE-GATED (removal is the canonical path). https://claude.ai/code/session_017GFLBnDy23AWBqvkbHHC41
AdaWorldAPI
changed the title
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 4
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In @.claude/board/AGENT_LOG.md:
- Line 3: Update the AGENT_LOG entry by replacing the placeholder "Commit: in
PR." with the concrete git commit SHA for the recorded run; locate the entry in
.claude/board/AGENT_LOG.md (the line containing "Commit: in PR.") and stamp the
full commit hash, and ensure the same entry also explicitly lists D-ids, the
tests run, and the outcome so the record is immutable and auditable.
In @.claude/board/EPIPHANIES.md:
- Around line 51-53: The cross-reference line contains an unescaped leading hash
(`#487`) that is being parsed as an ATX heading; edit the cross-ref so `#487` is
not at the start of its line — either move it to the same line as the preceding
"PR" token (e.g., "PR `#487`") or escape the hash (`\`#487``) so the markdown linter
(MD018) no longer treats it as a heading.
In @.claude/board/TECH_DEBT.md:
- Around line 939-941: The file contains conflicting status lines for the
CollapseGateEmission debt row: one line says "Status: CLOSED 2026-06-11 (moot)"
and later it still has "**Status:** Open"; remove or update the
redundant/incorrect line so only a single authoritative status remains for the
CollapseGateEmission entry (keep the CLOSED line per the PR `#477` note and delete
the "**Status:** Open" line or replace it with matching closed status).
In `@crates/cognitive-shader-driver/src/mailbox_soa.rs`:
- Around line 129-133: The sentinel collision occurs because last_active_cycle
(the array named last_active_cycle) uses u32::MAX as "never consumed" while
current_cycle wraps with wrapping_add(1); fix by changing last_active_cycle from
[u32; N] with u32::MAX sentinel to [Option<u32>; N] initialized to None, update
all checks/assignments to use Some(cycle) / None (e.g., replace comparisons
against u32::MAX and direct assignments with Some(current_cycle)), and update
any logic in the same module referencing current_cycle and last_active_cycle
(including the other occurrence around the 223-225 region) to handle Option<u32>
accordingly so a wrapped current_cycle will never collide with the "never
consumed" state.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro Plus
Run ID: eaf6f32a-1384-4477-93b0-0bd342bdbe51
📒 Files selected for processing (12)
.claude/board/AGENT_LOG.md.claude/board/EPIPHANIES.md.claude/board/LATEST_STATE.md.claude/board/TECH_DEBT.md.claude/plans/cycle-coherent-soa-snapshot-v1.mdCLAUDE.mdcrates/cognitive-shader-driver/src/mailbox_soa.rscrates/lance-graph-contract/src/collapse_gate.rscrates/lance-graph-contract/src/episodic_edges.rscrates/lance-graph-contract/src/kanban.rscrates/lance-graph-contract/src/lib.rscrates/lance-graph/src/graph/witness_tombstone.rs
| @@ -1,3 +1,7 @@ | |||
| ## 2026-06-11 — tombstone commit: emission artifacts removed per PR #477 follow-up | |||
|
|
|||
| **Main thread (Fable, session splat3d-cpu-simd-renderer).** Executed the PR #477 documented follow-up (the "what does NOT exist" table → source reality): removed `CollapseGateEmission` from `lance-graph-contract::collapse_gate` (+ lib.rs re-export; `MailboxId`/`MergeMode`/`GateDecision` survive), removed `MailboxSoA::emit()`, renamed `last_emission_cycle` → `last_active_cycle`, added in-place `consume_firing(row)` successor (same threshold + same-cycle-idempotency semantics, no carrier object), reworded 4 stale doc references (kanban/episodic_edges/witness_tombstone/mailbox_soa header), superseded the CLAUDE.md Baton-scoping block, fixed cycle-coherent-soa-snapshot-v1 D-SOA-SNAP-1/2 to generic `SnapshotProvider::Column` (closes #477 CodeRabbit Critical — contract stays zero-dep), closed TD-COLLAPSE-GATE-SMALLVEC-1 as moot. Verified #477 codex P2 (`verify_layout` ColumnOutOfBounds) already fixed on main with regression test. Tests: contract 594 (−8 emission, +2 gate/merge), driver 85 (emit tests → consume tests, +1 OOB), clippy clean, workspace check clean. Commit: in PR. | |||
There was a problem hiding this comment.
Record the exact commit SHA in this AGENT_LOG entry.
Commit: in PR. is not immutable/auditable for run traceability. Please stamp the concrete commit hash in this prepended record.
As per coding guidelines, .claude/board/AGENT_LOG.md entries for completed runs must include D-ids, commit, tests, and outcome in the same commit.
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In @.claude/board/AGENT_LOG.md at line 3, Update the AGENT_LOG entry by
replacing the placeholder "Commit: in PR." with the concrete git commit SHA for
the recorded run; locate the entry in .claude/board/AGENT_LOG.md (the line
containing "Commit: in PR.") and stamp the full commit hash, and ensure the same
entry also explicitly lists D-ids, the tests run, and the outcome so the record
is immutable and auditable.
Source: Coding guidelines
Comment on lines
+51
to
+53
| **Cross-ref:** PR #477 (three-tier model + "what does NOT exist" table); PR | ||
| #487 (tombstone commit — emission artifacts removed; `consume_firing` is the | ||
| in-place ownership successor); `docs/architecture/soa-three-tier-model.md` |
There was a problem hiding this comment.
Keep #487 off its own line.
#487 at the start of Line 52 is parsed as an ATX heading (MD018), which breaks the cross-reference and trips markdownlint. Keep it on the same line as PR or escape the hash.
Fix
- **Cross-ref:** PR `#477` (three-tier model + "what does NOT exist" table); PR
- `#487` (tombstone commit — emission artifacts removed; `consume_firing` is the
+ **Cross-ref:** PR `#477` (three-tier model + "what does NOT exist" table); PR `#487` (tombstone commit — emission artifacts removed; `consume_firing` is the📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
Suggested change
| **Cross-ref:** PR #477 (three-tier model + "what does NOT exist" table); PR | |
| #487 (tombstone commit — emission artifacts removed; `consume_firing` is the | |
| in-place ownership successor); `docs/architecture/soa-three-tier-model.md` | |
| **Cross-ref:** PR `#477` (three-tier model + "what does NOT exist" table); PR `#487` (tombstone commit — emission artifacts removed; `consume_firing` is the | |
| in-place ownership successor); `docs/architecture/soa-three-tier-model.md` |
🧰 Tools
🪛 markdownlint-cli2 (0.22.1)
[warning] 52-52: No space after hash on atx style heading
(MD018, no-missing-space-atx)
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In @.claude/board/EPIPHANIES.md around lines 51 - 53, The cross-reference line
contains an unescaped leading hash (`#487`) that is being parsed as an ATX
heading; edit the cross-ref so `#487` is not at the start of its line — either
move it to the same line as the preceding "PR" token (e.g., "PR `#487`") or escape
the hash (`\`#487``) so the markdown linter (MD018) no longer treats it as a
heading.
Source: Linters/SAST tools
Comment on lines
+939
to
941
| **Status: CLOSED 2026-06-11 (moot)** — `CollapseGateEmission` was removed entirely per the PR #477 three-tier model (zero-copy SoA, no inter-mailbox handoff type; tombstone commit). No carrier, no Vec, nothing to optimize. | ||
|
|
||
| **Status:** Open |
There was a problem hiding this comment.
Resolve conflicting status lines for this debt row.
Line 939 marks this row as closed, but Line 941 still says **Status:** Open. Keep a single authoritative status line to avoid conflicting board state.
🛠️ Proposed fix
-**Status: CLOSED 2026-06-11 (moot)** — `CollapseGateEmission` was removed entirely per the PR `#477` three-tier model (zero-copy SoA, no inter-mailbox handoff type; tombstone commit). No carrier, no Vec, nothing to optimize.
-
-**Status:** Open
+**Status:** CLOSED 2026-06-11 (moot) — `CollapseGateEmission` was removed entirely per the PR `#477` three-tier model (zero-copy SoA, no inter-mailbox handoff type; tombstone commit). No carrier, no Vec, nothing to optimize.As per coding guidelines, governance files should keep status updates authoritative on the status line.
📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
Suggested change
| **Status: CLOSED 2026-06-11 (moot)** — `CollapseGateEmission` was removed entirely per the PR #477 three-tier model (zero-copy SoA, no inter-mailbox handoff type; tombstone commit). No carrier, no Vec, nothing to optimize. | |
| **Status:** Open | |
| **Status:** CLOSED 2026-06-11 (moot) — `CollapseGateEmission` was removed entirely per the PR `#477` three-tier model (zero-copy SoA, no inter-mailbox handoff type; tombstone commit). No carrier, no Vec, nothing to optimize. |
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In @.claude/board/TECH_DEBT.md around lines 939 - 941, The file contains
conflicting status lines for the CollapseGateEmission debt row: one line says
"Status: CLOSED 2026-06-11 (moot)" and later it still has "**Status:** Open";
remove or update the redundant/incorrect line so only a single authoritative
status remains for the CollapseGateEmission entry (keep the CLOSED line per the
PR `#477` note and delete the "**Status:** Open" line or replace it with matching
closed status).
Source: Coding guidelines
Comment on lines
+129
to
+133
| // u32::MAX is the "never consumed" sentinel. current_cycle starts at 0 | ||
| // and advances via wrapping_add(1); in practice it never reaches | ||
| // u32::MAX during a session, so the first emit on any cycle is always | ||
| // permitted (u32::MAX != any valid cycle stamp). | ||
| last_emission_cycle: [u32::MAX; N], | ||
| // u32::MAX during a session, so first consumption on any cycle is | ||
| // always permitted (u32::MAX != any valid cycle stamp). | ||
| last_active_cycle: [u32::MAX; N], |
There was a problem hiding this comment.
Prevent sentinel collision at cycle wraparound.
last_active_cycle uses u32::MAX as the “never consumed” sentinel, but current_cycle wraps with wrapping_add(1). When cycle reaches u32::MAX, untouched rows are incorrectly treated as already-consumed for that full cycle.
💡 Suggested fix
- // u32::MAX is the "never consumed" sentinel. current_cycle starts at 0
- // and advances via wrapping_add(1); in practice it never reaches
- // u32::MAX during a session, so first consumption on any cycle is
- // always permitted (u32::MAX != any valid cycle stamp).
+ // u32::MAX is the "never consumed" sentinel. `tick()` reserves this value
+ // so `current_cycle` never equals the sentinel.
last_active_cycle: [u32::MAX; N],
@@
- pub fn tick(&mut self) {
- self.current_cycle = self.current_cycle.wrapping_add(1);
- }
+ pub fn tick(&mut self) {
+ // Reserve u32::MAX for last_active_cycle sentinel.
+ self.current_cycle = if self.current_cycle == u32::MAX - 1 {
+ 0
+ } else {
+ self.current_cycle + 1
+ };
+ }Also applies to: 223-225
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@crates/cognitive-shader-driver/src/mailbox_soa.rs` around lines 129 - 133,
The sentinel collision occurs because last_active_cycle (the array named
last_active_cycle) uses u32::MAX as "never consumed" while current_cycle wraps
with wrapping_add(1); fix by changing last_active_cycle from [u32; N] with
u32::MAX sentinel to [Option<u32>; N] initialized to None, update all
checks/assignments to use Some(cycle) / None (e.g., replace comparisons against
u32::MAX and direct assignments with Some(current_cycle)), and update any logic
in the same module referencing current_cycle and last_active_cycle (including
the other occurrence around the 223-225 region) to handle Option<u32>
accordingly so a wrapped current_cycle will never collide with the "never
consumed" state.
…conflicts Board prepend zones (AGENT_LOG, EPIPHANIES) auto-merged; no semantic overlap — main-side additions (#485 helix probe, #486 docs) reference none of the tombstoned emission identifiers (verified by grep: only intentional historical mentions in comments remain). https://claude.ai/code/session_01PBTGaPCSnnt6u3pjXpbLwY
AdaWorldAPI
pushed a commit
that referenced
this pull request
Jun 13, 2026
Scope: harden the SurrealDB kv-lance backend so it's production-grade for the VIEW path (external SurrealQL surface over the leading Lance dataset). Does NOT touch the production cognitive-state writer — per the 2026-05-28 ruling (E-RUBICON-RACTOR) and the 2026-06-09 polyglot plan, LanceDB leads as the in-binary writer; SurrealDB is a view/dialect, and lance-graph commits direct. Key inventory finding pulled from polyglot §2.2: kv-lance is 18/19 implemented in-tree (get/keys/keysr/scan/scanr + writes + savepoints, MVCC via Lance versions, Timeline time-travel, background optimizer, ~6 k internal tests). The surreal_container BLOCKED(C/D) marker is stale — work is hardening + closing named gaps, not building. 8 deliverables D-KVL-1..8: - D-KVL-1: lance =7.0.0 / lancedb =0.30.0 pin discipline (in-binary consumer = no external pressure to chase upstream); CI assertion - D-KVL-2 (P0): Gap #2 fix — replace the LWW MergeInsert relic in flusher.rs:251 with MergeMode-tagged dispatch (recommended) or hard-partitioning proof; regression test for silent-drop - D-KVL-3: WAL — ciborium CBOR -> raw #[repr(C)]-LE + CRC32C + WAL-version byte (per I-LEGACY-API-FEATURE-GATED) - D-KVL-4: consumer-side crash-recovery suite (WAL-fsync / memtable-flush / Lance-commit / Lance-GC kill phases) - D-KVL-5: 1-hour multi-writer + multi-reader soak; budget <0.01% - D-KVL-6: schema-evolution boundary doc + tests under the pin - D-KVL-7: surreal_container unblock = polyglot D-PG-6; kanban VIEW smoke test against Tier-A direct-Lance writer (the contract test) - D-KVL-8: doctrine record — EPIPHANIES E-KVLANCE-VIEW-NOT-WRITER extends #487's E-OUTER-BOUNDARY-IS-ORM-1 over the four-tier map (A: in-binary = lance-graph direct; B-view: SurrealDB on hardened kv-lance; B-HA: kv-tikv; B-fallback: kv-rocksdb) Out of scope: kv-rocksdb hardening (SurrealDB-upstream tested); kv-tikv HA (separate plan); lance-graph Tier-A writer improvements; chasing lance/lancedb past the pin. This commit lands the plan file + INTEGRATION_PLANS prepend + STATUS_BOARD section (Board-Hygiene Rule). EPIPHANIES E-KVLANCE-* + TECH_DEBT TD-KVLANCE-LWW-RELIC wording pending user review. https://claude.ai/code/session_01PBTGaPCSnnt6u3pjXpbLwY
AdaWorldAPI
pushed a commit
that referenced
this pull request
Jun 13, 2026
Scope: harden the SurrealDB kv-lance backend so it's production-grade for the VIEW path (external SurrealQL surface over the leading Lance dataset). Does NOT touch the production cognitive-state writer — per the 2026-05-28 ruling (E-RUBICON-RACTOR) and the 2026-06-09 polyglot plan, LanceDB leads as the in-binary writer; SurrealDB is a view/dialect, and lance-graph commits direct. Key inventory finding pulled from polyglot §2.2: kv-lance is 18/19 implemented in-tree (get/keys/keysr/scan/scanr + writes + savepoints, MVCC via Lance versions, Timeline time-travel, background optimizer, ~6 k internal tests). The surreal_container BLOCKED(C/D) marker is stale — work is hardening + closing named gaps, not building. 8 deliverables D-KVL-1..8: - D-KVL-1: lance =7.0.0 / lancedb =0.30.0 pin discipline (in-binary consumer = no external pressure to chase upstream); CI assertion - D-KVL-2 (P0): Gap #2 fix — replace the LWW MergeInsert relic in flusher.rs:251 with MergeMode-tagged dispatch (recommended) or hard-partitioning proof; regression test for silent-drop - D-KVL-3: WAL — ciborium CBOR -> raw #[repr(C)]-LE + CRC32C + WAL-version byte (per I-LEGACY-API-FEATURE-GATED) - D-KVL-4: consumer-side crash-recovery suite (WAL-fsync / memtable-flush / Lance-commit / Lance-GC kill phases) - D-KVL-5: 1-hour multi-writer + multi-reader soak; budget <0.01% - D-KVL-6: schema-evolution boundary doc + tests under the pin - D-KVL-7: surreal_container unblock = polyglot D-PG-6; kanban VIEW smoke test against Tier-A direct-Lance writer (the contract test) - D-KVL-8: doctrine record — EPIPHANIES E-KVLANCE-VIEW-NOT-WRITER extends #487's E-OUTER-BOUNDARY-IS-ORM-1 over the four-tier map (A: in-binary = lance-graph direct; B-view: SurrealDB on hardened kv-lance; B-HA: kv-tikv; B-fallback: kv-rocksdb) Out of scope: kv-rocksdb hardening (SurrealDB-upstream tested); kv-tikv HA (separate plan); lance-graph Tier-A writer improvements; chasing lance/lancedb past the pin. This commit lands the plan file + INTEGRATION_PLANS prepend + STATUS_BOARD section (Board-Hygiene Rule). EPIPHANIES E-KVLANCE-* + TECH_DEBT TD-KVLANCE-LWW-RELIC wording pending user review. https://claude.ai/code/session_01PBTGaPCSnnt6u3pjXpbLwY
AdaWorldAPI
changed the title
CollapseGateEmission + MailboxSoA::emit(); in-place consume_firing() successor
Merged
4 tasks
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What
PR #477's "what does NOT exist (and must not be reintroduced)" table, made source-true. #477 documented the target state; this PR removes the superseded emission artifacts from source.
Removed
CollapseGateEmission(contract::collapse_gate)MailboxId/MergeMode/GateDecisionsurvive — addressing / merge / gate concepts, not emission concepts. A dated removal marker explains the lineage at the old site.MailboxSoA::emit()consume_firing(row) -> bool— same threshold + same-cycle-idempotency semantics, zero carrier object. The owner reads the row's columns in place (edge / qualia / meta / entity_type), then stamps. This is the ownership-transfer shape: no DTO, the SoA itself is the active record.last_emission_cyclelast_active_cycle(in-place consumption stamp), per the #477 follow-up list verbatim.wire_cost_bytes() = 13 + 10·baton_countDoc reality sync (same commit)
docs/architecture/soa-three-tier-model.md):kanban.rs(R4 witness-as-pointer ×2),episodic_edges.rs(LE byte grammar consumers),witness_tombstone.rs(E-BATON-1 scoping),mailbox_soa.rsmodule header.cycle-coherent-soa-snapshot-v1.mdD-SOA-SNAP-1/2 fixed to genericMailboxSoaSnapshot<C>+SnapshotProvider::Column— closes the SoA envelope: zero-copy register-file model + LE contract + ontology resolution audit #477 CodeRabbit Critical (the original sketch putVec<Arc<MultiLaneColumn>>in contract, which would have created a contract → ndarray dependency). Contract stays zero-dep; lance-graph bindsColumn = MultiLaneColumn; same triangulation asSoaEnvelope.Verified, not re-fixed
The #477 codex P2 on
verify_layout()(column end-offset past stride passing the summed-width check) is already fixed on main —ColumnOutOfBoundswithchecked_addoverflow guard + regression testcolumn_past_stride_caughtusing codex's exact example. The GitHub thread was never marked resolved; the code is correct. No change needed.Board hygiene (same commit)
LATEST_STATE.mdcontract inventory:CollapseGateEmissionmarked REMOVED; TD row updated.TECH_DEBT.md: TD-COLLAPSE-GATE-SMALLVEC-1 closed as moot (no carrier, no Vec, nothing to optimize).AGENT_LOG.md: entry prepended.Gates
cargo test -p lance-graph-contract --libcargo test -p cognitive-shader-driver --libcargo clippy -p lance-graph-contract -p cognitive-shader-driver --all-targetscargo check --workspaceNot in scope
OntologyRegistrylinear-scan → O(1) HHTL lookup (separate SoA envelope: zero-copy register-file model + LE contract + ontology resolution audit #477 follow-up).CausalEdge64types (causal-edge vs thinking-engine) — needs its own disambiguation PR.