plan(sprint-13/preflight): Wave H — 13 planners + W-Meta-Opus + 4-agent quality lifecycle#391
Merged
Merged
Conversation
PP-3 Opus planner delivered the par_* rayon streams spec (`pr-sprint-13-rayon-streams.md`, 673 lines / 29.6 KB) for the sprint-12 D-CSV-11 forward-iter follow-on. Key findings from PP-3's recon: - ndarray's rayon feature is named `rayon` (not `parallel`) per `/home/user/ndarray/Cargo.toml:148`; spec re-uses existing flag - docs.rs metadata at line 255 already includes `"rayon"` in features so par_* doc comments will render on docs.rs - existing `par_rayon` bench precedent at line 122 confirms spelling - scalar scaffolds confirm 8/8/16 B row sizes with repr(C, align(N)) supporting the cache-line chunk table (8/8/4 rows per 64 B line) Sections §0-§12: scope, trait surface (IndexedParallelIterator mandatory for enumerate/zip/with_min_len), per-stream sketch + mod.rs re-export, feature gate, cache-line chunk strategy, determinism contract (with VSA-bundling iron rule cross-ref), 18 tests, risk matrix, cross-refs, LOC ~270 actual, acceptance checklist, 3 OQs. PP-3 chose D-CSV-17 for the deliverable id; main thread will reconcile numbering when all 12 PP outputs aggregate (PP-3 noted 'D-CSV-16 reserved for PP-2 splat on-Think methods' but per main thread's brief PP-4 owns on-Think methods + PP-5 owns WitnessIndexCamPq — numbering will be resolved in the W-Meta-Opus pass). 11 more Opus planners + W-Meta-Opus pending. https://claude.ai/code/session_01UwJuKqP828qyX1VkLgGJFS
PP-1 (Opus): `.claude/plans/cognitive-substrate-convergence-v3.md` — sprint-13 execution plan + sprint-12 outcomes roll-up, mirroring v2's section numbering with explicit `🆕 v3 (sprint-13)` markers. PP-8 (Opus): `.claude/agents/worker-template-v2.md` — closes CSI-13 (workers now own full integration: lib.rs touches, Cargo.toml deps, workspace member additions when creating new crates, field-isolation matrix tests at layout-bit boundaries per I-LEGACY-API-FEATURE-GATED). Sprint-13 worker spawn template; copy-paste skeleton with mandatory sections (workspace + branch + worker ID + mandatory reads + scope discipline + integration responsibility + validation + output format + forbidden actions + codex P1 anticipation). 3 of 12 PP planners done (PP-1, PP-2, PP-3, PP-8). 9 still in flight + W-Meta-Opus. https://claude.ai/code/session_01UwJuKqP828qyX1VkLgGJFS
…sprint-12 meta + board refresh + OQ catalog Five more Opus planners landed: PP-4 (Opus): `pr-sprint-13-think-methods.md` (674 LOC) — D-CSV-14 spec for splat-ops on Think carrier. Found: zero external callers of the 4 free fns (deprecation cycle is conservative not forcing); Think struct doesn't exist yet (sprint-13 introduces minimum-viable with splat_field + cycle; sprint-15+ accretes trajectory/awareness/etc). PP-5 (Opus): `pr-sprint-13-witness-cam-pq.md` — D-CSV-16 spec for real ndarray CAM-PQ wiring replacing the Wave G HashMap placeholder. PP-7 (Opus): `.claude/board/sprint-log-12/meta-review.md` — sprint-12 closing meta-review. PP-10 (Opus): LATEST_STATE.md (+50 lines) + STATUS_BOARD.md (+15 lines) refresh. 7 new "Recently Shipped" PR rows (#383-#389), 19+ new contract types in inventory, sprint-13 queued specs surfaced. Phase A/B/C marked Shipped; Phase D partial; Phase E (sprint-13) section added. PP-11 (Opus): `.claude/board/sprint-log-13/oq-catalog.md` — 13 OQs (OQ-CSV-7..19); 10 block sprint-13 spawn; 3 sprint-14+ tracking (ndarray PR #116 pin/vendor, VAMPE+Jirak scope, splat persistence). 7 of 12 PP planners done. 5 still in flight (PP-2 iron-rules done separately; PP-6 SIMD spec, PP-9 PR_ARC backfill, PP-12 cross-repo audit pending). https://claude.ai/code/session_01UwJuKqP828qyX1VkLgGJFS
PP-6 (Opus) — pr-sprint-13-simd-i4.md (982 LOC, 46 KB). All 13 sections.
Scope: 5 batch functions × {AVX-512, NEON} + scalar fallback for D-CSV-13b.
AVX-2 / WASM / VNNI / AMX explicitly out of scope for this PR.
i4 lane → register mapping: 8 qualia/AVX-512 vec on x86, 2 qualia/NEON on
ARM. Full intrinsic code shown for §3.1 dk_position + §3.2 trust_texture
(the 2 fns with #[repr(u8)] outputs). §3.3-§3.10 are structural sketches
with priority-order analysis.
Runtime dispatch via ndarray::simd_caps() (proven primitive at
ndarray/src/hpc/simd_caps.rs:98). Falls back to scalar (the W-G3 batch
API) when target_feature missing.
Semantic-equivalence iron rule citation: I-LEGACY-API-FEATURE-GATED applies
DIRECTLY — same function name MUST produce byte-identical output across
SIMD/scalar feature gates. 10 equivalence tests; exhaustive for the 2 fns
with bounded discriminant space.
Key findings from PP-6 recon:
- W-G3 scalar batch API at mul.rs:627-695 (commit 03ce219) — the target API
- _mm512_cmpgt_epi8_mask (simd_avx512.rs:1612) + _mm512_mask_blend_epi8
(L641) — proven reuse patterns
- GateDecision::Block { reason: String } payload blocks naive #[repr(u8)]
packing — spec carves out gate_decision_disc_batch discriminant-only
variant
- _mm512_mask_compressstoreu_epi8 needs AVX-512VBMI2 (not just F+BW);
spec downgrades to scalar byte-extract for portability (R-6 risk row)
5-PR slicing plan: W-SIMD-1..W-SIMD-5 (per-fn rollout).
Per-fn bench targets: ≥4× speedup on AVX-512 (≥2.5× for mul_assess).
11 of 13 PP planners done. PP-9 (PR_ARC backfill), PP-12 (cross-repo
audit), PP-13 (brutally-honest-tester) still in flight.
https://claude.ai/code/session_01UwJuKqP828qyX1VkLgGJFS
PP-12 (Opus): `.claude/knowledge/cross-repo-dependency-audit-2026-05.md` — audits the sibling-repo coordination requirements before sprint-13 spawn. Topics: ndarray dep map (PR #147 status, cam_pq shape, rayon feature gate, simd_caps singleton, PR #116 hpc-extras), p64 / p64-bridge state (convergence drift), sprint-13 D-CSV-* coordination order, and a cross-repo PR coordination protocol proposal. 12 of 13 PP planners done. PP-9 (PR_ARC backfill) + PP-13 (brutally- honest-tester) still in flight. https://claude.ai/code/session_01UwJuKqP828qyX1VkLgGJFS
…ntries (#383..#390) PP-9 (Opus) prepended 8 PR entries to PR_ARC_INVENTORY.md (1537→1903 lines, +366 / +50,619 bytes). APPEND-ONLY rule respected — no prior entries touched. Top→bottom reverse chronological: #390 sprint-12/wave-G (In PR, `bad0875`) — grade A−. D-CSV-5b cutover + D-CSV-6b WitnessCorpus + D-CSV-13 batch + D-CSV-15 Jirak math. #389 sprint-12 wave-F codex P2 follow-up — AttentionMaskBackend impl + canonical MailboxId (CSI-10). #388 sprint-12 Wave F fleet (12 Sonnet + 1 Opus) — grade B. D-CSV- 10/11/12 scaffolds + AttentionMask + plan v2. #387 sprint-11 Wave E — grade A−. D-CSV-8 MUL i4 scalar + D-CSV-9 8-channel transcoder (Option R-3). #386 sprint-11 Wave D — grade B+. D-CSV-7 MailboxSoA + D-CSV-6a WitnessCorpus core. #385 sprint-11 Wave C — grade B+. D-CSV-5a sibling QualiaI4Column. #384 sprint-11 Wave B — grade A. D-CSV-2 QualiaI4_16D + OQ-CSV-1 Option α ratification. #383 sprint-11 Wave A — grade A−. D-CSV-1 v2 layout + D-CSV-3 signed mantissa + D-CSV-4 CollapseGateEmission. Every entry mirrors the #381/#379 template — Header / Confidence / Added / Locked / Deferred / Docs / Cross-refs — with forward+backward linkages (#383→#385→#390 D-CSV-5 chain; E-META-10 catch in #383 → iron-rule promotion in #390 W-G5; etc.). Total PR headers in file: 28 (was 20). 13 of 14 planners done. Only PP-13 (brutally-honest-tester agent) still in flight. https://claude.ai/code/session_01UwJuKqP828qyX1VkLgGJFS
PP-13 (Opus) — `.claude/agents/brutally-honest-tester.md` (708 LOC, 29.6 KB). User-requested agent that bundles the stable-Rust toolchain into a strict pre-merge gate plus codex-style P1 anticipation. Structure: - YAML frontmatter (model: opus; tools: Read/Glob/Grep/Bash) - Mandatory reads Tier 0/1/2 (iron-rules-doctrine PP-2 + sprint-log meta-reviews + worker-template-v2 §5.5/§9 + diff-triggered domain knowledge docs) - §1 Toolchain across 5 tiers: Mandatory (clippy + fmt + audit + deny), Strict (clippy pedantic/nursery), Recommended (machete + geiger + semver-checks + outdated), Targeted (spellcheck + public-api + bloat + nextest), Formal-ish (kani + loom). Explicit non-fit: Miri / fuzz / mutants (nightly or heavy) - §2 Codex P1 catalogue AP1..AP8 (v1-under-v2-alias / bit-collision / sub-crate [workspace] / lib.rs orphan / cross-repo mod.rs orphan / new-abstraction / unsafe-without-SAFETY / REST-endpoint-drift). Each pattern documents grep targets + rule + suggested fix + sprint instance citation (CSI-2/7/8/9 etc.) - §3 Output format — markdown verdict block with P0/P1/P2 lists + toolchain rollup + anti-pattern rollup + ternary LAND/HOLD/REJECT - §4 Workflow integration — CCA2A slot post-impl pre-meta-review; draft `.github/workflows/honest-tester.yml` (flagged as D-CSV-18) - §5 BOOT.md Tier-1 trigger row - §6 Promotion ceremony (3 gates + 1 false-positive + 1 missed-bug) - §7 Cross-refs - §8 "One sentence that should survive any refactor" + authorship Mirrors `truth-architect.md` + `certification-officer.md` format precedent. Cites sprint-11 meta-review-opus CSI-2 (the 4-instance PR #383 cluster that drove I-LEGACY-API-FEATURE-GATED promotion), sprint-12 CSI-7/8/9 (orphan-pattern that motivated worker-template-v2), worker-template-v2 §5.5+§9 (the IMPL-side discipline this REVIEW agent mirrors). ALL 14 PP planners complete (12 original + PP-2 iron-rules-doctrine ran early + PP-13 brutally-honest-tester added mid-fleet per user request). W-Meta-Opus honest review dispatch next. https://claude.ai/code/session_01UwJuKqP828qyX1VkLgGJFS
…patterns knowledge doc
Two preflight closeout artifacts:
1. .claude/board/sprint-log-13/preflight-meta-review-opus.md (W-Meta-Opus, ~196 LOC)
- Wave H grade: B+ (depressed by plan-vs-git drift in PP-1/PP-12 + ID/OQ-numbering
drift across PP-1..PP-6/PP-11; impl specs PP-3/4/5/6 + structural fixes
PP-2/8/13 are engineer-ready)
- Three notable CSIs surfaced: CSI-19 (D-CSV-/OQ-CSV-numbering drift),
CSI-20 (multiple planners drifted from main on PR #390-dependent state),
CSI-23 (PR #390 was the gate for sprint-13 spawn — now resolved with #390 merge)
- 8-item sprint-13 spawn checklist
- Per-planner verifications: PP-4 zero-callers-of-splat-fns VERIFIED TRUE,
PP-4 no-Think-struct VERIFIED TRUE, PP-5 CamPqIndexPlaceholder citation
VERIFIED ACCURATE, PP-9 APPEND-ONLY discipline VERIFIED, PP-12 false
ndarray-PR-#147-merged claim VERIFIED FALSE
2. .claude/knowledge/codex-p1-anti-patterns.md (~455 LOC, PP-13 companion)
- Operational catalogue the brutally-honest-tester (PP-13) loads at Tier-1
- §1 codex bot's review shape (P1/P2/P3 frequencies sprint-11/12)
- §2 anti-pattern catalogue AP1-AP8 with grep targets + rules + fixes
+ sprint-11/12 instance citations (commits 42b3215, 33509ab, d4e5bbc, 2a1a1e3)
- §3 stable-Rust toolchain across 5 tiers (Mandatory/Strict/Recommended/
Targeted/Formal-ish; explicit non-fit: Miri/cargo-careful/cargo-fuzz)
- §4 severity convention P0/P1/P2/P3 → LAND/HOLD/REJECT
- §5 workflow integration with CCA2A loop diagram
- §6 APPEND-ONLY maintenance protocol with promotion track to iron rules
- §8 'one sentence that should survive any refactor' anchor
Closes the 'create .claude/agents + knowledge for the new codex savant' request:
PP-13 created the agent card; this commit ships the companion knowledge doc.
…itect / baton-handoff-auditor / preflight-drift-auditor
Completes the four-agent quality lifecycle alongside PP-13 brutally-honest-tester:
PRE-PLAN → convergence-architect (PP-14, divergent expansion)
PRE-SPAWN → preflight-drift-auditor (PP-16, automates W-Meta-Opus audit)
DURING-IMPL → baton-handoff-auditor (PP-15, cross-crate boundary mismatches)
POST-IMPL → brutally-honest-tester (PP-13, codex-class within-crate bugs)
PP-14 convergence-architect (444 + 719 LOC):
- Creative-genius "0-friction boundary" hunter; verdict OPPORTUNITY/
WORTH-EXPLORING/DROP (never REJECT — psychological safety for ideation)
- EP1-EP8 expansion-pattern catalogue: 0-friction baton handover,
algebraic-operator reuse, role-keyed identity migration,
carrier-as-grammar, iron-rule axis prediction, sprint-N-bug-becomes-
feature, cross-repo synergy without API change, doctrine concentration
- Tools: + WebSearch, WebFetch, Agent, AskUserQuestion, ToolSearch,
HF paper_search/hf_doc_search/hf_doc_fetch/hf_hub_query (cross-pollination
from external papers is the role; Jirak 2016 precedent → I-NOISE-FLOOR-JIRAK)
PP-15 baton-handoff-auditor (502 + 897 LOC):
- Brutal cross-boundary reviewer; verdict CATCH-CRITICAL/CATCH-LATENT/CLEAN
- BAP1-BAP10 boundary anti-pattern catalogue: DTO field-shape silent
drift, rename-without-downstream-sweep (CSI-15), lib.rs/mod.rs orphan
(CSI-8/9), Cargo.toml [workspace] self-declaration (CSI-7), REST without
OrchestrationBridge, sprint-N TD-* consumed as resolved (CSI-9/20),
D-id/OQ-id numbering collision (CSI-19), iron-rule carrier-boundary
violation, feature-gated v1-API alias (I-LEGACY-API-FEATURE-GATED
precedent), Producer<Option<X>>/Consumer-unwrap tail panic
- Commit SHAs cited: 42b3215, b44ce87, d4e5bbc, 0ae9f90, 2a1a1e38,
2a3885d2, 04620aa, b526485
- Tools: + ToolSearch, mcp__github__{get_file_contents, pull_request_read,
list_pull_requests, get_commit, search_code} (targeted cross-repo
single-file reads per CLAUDE.md zipball-for-bulk-MCP-for-targeted rule)
PP-16 preflight-drift-auditor (474 + 686 LOC):
- Pre-spawn drift hunter that automates the W-Meta-Opus manual audit
method; verdict SPAWN-BLOCKED/SPAWN-CAUTION/SPAWN-CLEAR
- PD1-PD10 catalogue: stale-CSI-resolved-via-commit (CSI-9), cross-planner
D-id collision (CSI-19), pending-iron-rule-as-canonical (CSI-20),
false cross-repo PR-merge (PP-12 ndarray PR #147), renamed-symbol
downstream drift (CSI-15), spec-internal contradiction, iron-rule-on-
wrong-branch, missing-integration-responsibility (CSI-13/8),
TD-placeholder-consumed-as-resolved, plan-vs-board-inventory drift
- Each PD has copy-paste-runnable verification command (git log / grep /
file:line); W-Meta-Opus is cited as primary source 16+31 occurrences
- Tools: + ToolSearch, mcp__github__{get_file_contents, list_pull_requests,
get_commit, pull_request_read, list_commits, search_pull_requests}
(cross-repo PR-merge claim verification is the role — exactly the gap
that produced PP-12's false ndarray-PR-#147 drift)
All three cards: model: opus (multi-source synthesis per Model Policy),
read-only tools (no Edit/Write/MultiEdit — savants are auditors; writes
happen on main thread after their verdicts).
Both PP-14 and PP-15 independently flagged a meta-finding during prior-art
reads: their own sibling docs (PP-13 brutally-honest-tester +
codex-p1-anti-patterns) treat I-LEGACY-API-FEATURE-GATED as canonical
when it is still pending PR #390 ratification. Captured as BAP6 / PD3
respectively; aggregation sweep deferred to sprint-13 W-G7 cleanup.
…t scope
User clarification: "tools" = Rust toolchain commands per agent, not Claude
tool YAML field. Each card now declares its targeted toolchain subsection
keyed to its role, cross-referencing the comprehensive canonical list owned
by PP-13 brutally-honest-tester.
PP-13 brutally-honest-tester (§1.5 formal-ish tier expanded):
+ cargo mutants (mutation testing; P2 informational unless deliverable gates it)
+ cargo tarpaulin (line coverage, Linux-only; P2 informational unless gated)
Aligns §1 with the user's canonical 3-tier list (Mandatory: clippy/fmt/audit/deny;
Quality: machete/geiger/semver-checks/spellcheck/public-api; Heavier-opt-in:
kani/loom/mutants/tarpaulin).
PP-14 convergence-architect (new §1.5 — targeted for divergent ideation):
- No Rust gate commands; pre-impl synthesis has nothing to lint
- Uses Read/Glob/Grep + WebSearch/WebFetch/HF paper_search for cross-pollination
- Optional cargo doc/tree/expand for inspecting existing algebra surface
- Explicit non-use of clippy/fmt/audit/deny/test/bench
- Cross-ref to PP-13 §1 for the comprehensive canonical list
PP-15 baton-handoff-auditor (new §1.5 — targeted for cross-boundary audit):
- cargo check --workspace --all-targets --all-features (B2/B6/B7 orphan/workspace)
- cargo tree -p <crate> --workspace (B2/B8 cross-crate + cross-repo dep paths)
- cargo metadata | jq '.workspace_members' (B7 [workspace] trap, CSI-7 precedent)
- cargo public-api --diff (B1/B2 DTO field-shape + cross-crate signature)
- cargo semver-checks check-release (B9 v1-under-v2 alias, I-LEGACY-API-FEATURE-GATED)
- cargo machete (B6 orphan signal; B2 stale dep)
- cargo expand -p <crate> (B1 macro-hidden shape drift)
- git log + mcp__github__get_commit (B3 sprint↔sprint, B6 false-merged claim)
- grep -rn <old-symbol> (B2 rename-without-downstream-sweep, CSI-15)
- Explicit non-use of within-crate clippy/fmt/audit/deny/kani/loom (PP-13 owns)
PP-16 preflight-drift-auditor (new §1.5 — pre-spawn plan/spec drift verification):
- git log master --oneline (Axis 1/4 stale-CSI + cross-repo PR-merge claims;
automates the PP-12 false ndarray-PR-#147 audit pattern)
- git log --all --grep '<CSI-N>' --oneline (Axis 1/3 commit-cited-for-CSI)
- git show <SHA> --stat (Axis 1/4/5 commit-touches-cited-files)
- git diff main...HEAD (Axis 3/6 pending-vs-canonical iron-rule citation)
- mcp__github__list_pull_requests + get_commit (Axis 4 cross-repo PR verification)
- grep -rhn 'D-CSV-[0-9]+' .claude/plans/ .claude/specs/ (Axis 2 D-id collision, CSI-19)
- grep -rhn 'OQ-CSV-[0-9]+' .claude/plans/ .claude/specs/ .claude/board/ (Axis 2 OQ-id)
- grep -n 'I-LEGACY-API-FEATURE-GATED' CLAUDE.md (Axis 3/6 iron-rule against main)
- grep -rn '<old-symbol>' .claude/specs/ .claude/plans/ .claude/agents/ (Axis 5)
- cargo tree -p <crate> (occasional; Axis 4 dep-shaped spec claim verification)
- Explicit non-use: pre-spawn agent runs before any code exists for cargo gates
Cross-references ensure no toolchain command is owned twice — PP-13 covers
within-crate gates; PP-15 covers cross-boundary cargo gates; PP-16 covers
pre-spawn git/grep drift verification; PP-14 covers no cargo gates at all
(divergent ideation has no diff to verify).
LOC after edits:
brutally-honest-tester: 708 → 713 (+5 mutants/tarpaulin rows)
convergence-architect: 444 → 471 (+27 §1.5 ideation toolchain)
baton-handoff-auditor: 502 → 532 (+30 §1.5 boundary-audit toolchain)
preflight-drift-auditor: 474 → 509 (+35 §1.5 pre-spawn drift toolchain)
…3/14/15/16 four-agent quality lifecycle Adds 4 trigger rows to the Knowledge Activation Protocol table (line 135+ in .claude/agents/BOOT.md) so the 4 new savant agents can be woken by trigger phrase without main-thread guesswork: PRE-PLAN → convergence-architect (divergent expansion ideation) PRE-SPAWN → preflight-drift-auditor (automates W-Meta-Opus audit) DURING-IMPL → baton-handoff-auditor (cross-boundary mismatch hunter) POST-IMPL → brutally-honest-tester (codex-class within-crate gate) Each row names: trigger phrase taxonomy + agent + mandatory knowledge docs to load BEFORE producing output (per BOOT.md Knowledge Activation discipline). Closes W-Meta-Opus spawn-checklist items 7 (PP-8 worker-template-v2 → canonical applies via worker-template usage) and 8 (PP-13 brutally-honest-tester → probation-tier via BOOT.md trigger row). The other three new savants (PP-14/ 15/16) are first-class members of the lifecycle from the start.
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: f70af37019
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
Comment on lines
+220
to
+221
| let shifted_up = _mm512_slli_epi16::<4>(nibble); | ||
| let coherence_i8 = _mm512_srai_epi16::<4>(shifted_up); |
There was a problem hiding this comment.
For AVX-512 inputs where a qualia dimension is negative (0x8..0xF), this sequence does not sign-extend the low nibble: shifting as i16 makes 0xF become 0x00F0, then arithmetic-right-shifting by 4 yields 0x000F rather than -1 because the sign bit is bit 15, not bit 7. The planned AVX-512 batch path will therefore treat all negative i4 dimensions as large positives and diverge from the scalar QualiaI4_16D::get semantics whenever thresholds depend on sign.
Useful? React with 👍 / 👎.
Comment on lines
+297
to
+299
| default = ["unity-catalog", "delta", "ndarray-hpc"] | ||
| # ... existing features ... | ||
| with-cam-pq = ["ndarray-hpc"] # implies ndarray-hpc; cannot be enabled without it. |
There was a problem hiding this comment.
As written, with-cam-pq = ["ndarray-hpc"] only makes with-cam-pq imply ndarray-hpc; Cargo features do not imply each other in reverse, so the default ndarray-hpc feature will not enable any #[cfg(feature = "with-cam-pq")] code. That contradicts §4.3/T11's expectation that plain cargo build -p lance-graph includes WitnessIndexCamPq, meaning the default build would silently omit the new index unless callers pass --features with-cam-pq explicitly.
Useful? React with 👍 / 👎.
Comment on lines
+366
to
+367
| let mut think = Think::from_field(std::mem::take(field), generation.saturating_sub(1)); | ||
| think.splat_gaussian(mean, variance, energy); |
There was a problem hiding this comment.
When an existing caller passes generation == 0, this shim seeds Think::cycle with 0 because of saturating_sub(1), then Think::splat_gaussian advances it to 1; the old free function would have stored generation 0 directly. Since generation is a public u32 parameter and the spec promises bit-identical forwarding for any input, this edge case changes observable SplatField::generation values and can break zero-based replay/test fixtures.
Useful? React with 👍 / 👎.
Merged
4 tasks
AdaWorldAPI
added a commit
that referenced
this pull request
May 16, 2026
…-post-merge plan(sprint-13/cleanup): CSI-19 D-id reconciliation + CSI-20 stale-claim sweep post PR #391
Merged
5 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.
Summary
Sprint-13 preflight planning fleet (Wave H) — 13 Opus planners + W-Meta-Opus honest review + the 4-agent quality lifecycle (codex-like savants) that closes the sprint-11/12 CSI-13 / CSI-18 / CSI-19 / CSI-20 process gaps. Pure planning + governance + agent-ensemble work; zero source-code touches.
Branched off main at
b526485(pre PR #390 merge); branch is currently 7 commits behind main (PR #390 only touchedcrates/, this PR only touches.claude/, so no merge conflicts on rebase — handled at merge time).What this PR delivers
Sprint-13 implementation specs (4, engineer-ready for Sonnet impl workers)
ndarray::simd_caps()) —.claude/specs/pr-sprint-13-simd-i4.mdThinkcarrier (zero external callers verified) —.claude/specs/pr-sprint-13-think-methods.mdndarray::hpc::cam_pq(Option A VSA-bind per I-VSA-IDENTITIES) —.claude/specs/pr-sprint-13-witness-cam-pq.md.claude/specs/pr-sprint-13-rayon-streams.mdSprint-13 plan v3 + sprint-12 close-out
.claude/plans/cognitive-substrate-convergence-v3.md— 19 sections, sprint-13 forward plan + sprint-12 outcomes roll-up.claude/board/sprint-log-12/meta-review.md— sprint-12 close (Wave F=B, Wave G=A−, sprint=B+).claude/board/sprint-log-13/oq-catalog.md— 13 OQs (OQ-CSV-7..19); 10 block sprint-13 spawn.claude/board/sprint-log-13/preflight-meta-review-opus.md— W-Meta-Opus honest review (Wave H grade B+)Doctrine + process
.claude/knowledge/iron-rules-doctrine.md— 4-axis meta-pattern consolidation (substrate-operator / statistical-model / data-semantics / API-version); promotion ceremony for future iron rules.claude/agents/worker-template-v2.md— closes CSI-13 by baking lib.rs/workspace registration into worker scope (§5 Integration responsibility + §9 Codex-P1 self-scan).claude/knowledge/cross-repo-dependency-audit-2026-05.md— runway audit.claude/board/PR_ARC_INVENTORY.mdbackfill — 8 PR entries prepended (impl(sprint-11/wave-A): D-CSV-1 + D-CSV-3 + D-CSV-4 — causal-edge v2 + signed-mantissa NARS + CollapseGateEmission #383..impl(sprint-12/wave-G): 6 Sonnet + 1 Opus — D-CSV-5b/6b/13/15 + E-META-10 iron rule (sprint-12 grade B+) #390).claude/board/LATEST_STATE.md+STATUS_BOARD.mdrefreshThe 4-agent quality lifecycle (codex-like savants)
convergence-architect(444 + 719 LOC, EP1-EP8 catalogue)preflight-drift-auditor(474 + 686 LOC, PD1-PD10 catalogue)baton-handoff-auditor(502 + 897 LOC, BAP1-BAP10 catalogue)brutally-honest-tester(713 + 455 LOC, AP1-AP8 catalogue)Each savant card declares a targeted §1.5 toolchain section keyed to its role with cross-references — no command is owned twice; cross-scope findings route forward via explicit "→ PP-N" pointers. BOOT.md Knowledge Activation table updated with 4 new trigger rows (one per savant).
Meta-findings from the W-Meta-Opus audit (kept in for transparency; sprint-13 W-G7 cleanup gates)
The W-Meta-Opus review (
sprint-log-13/preflight-meta-review-opus.md) surfaced three classes of drift in the planning fleet itself — captured as CSI-19 (cross-planner D-id/OQ-id collision), CSI-20 (stale-claim drift on PR #390-pending state in PP-1 §0.1 / PP-12), and a meta-finding that PP-13/PP-14/PP-15 citeI-LEGACY-API-FEATURE-GATEDas canonical when PR #390 had not yet merged at planner-write time. PR #390 has since merged so the citation is now correct retroactively. The remaining CSI-19 / CSI-20 reconciliation aggregation is deferred to sprint-13 W-G7 cleanup peroq-catalog.md.Test plan
wc -l)tools:field declares its tool surface (Read/Glob/Grep/Bash + WebSearch + HF paper_search for PP-14; mcp__github tools for PP-15/PP-16)sprint-log-13/preflight-meta-review-opus.md)codex-p1-anti-patterns.md)https://claude.ai/code/session_01UwJuKqP828qyX1VkLgGJFS
Generated by Claude Code