cognitive-shader-driver: emit CE64 from the SPOFC record (land #1294 on main)

[AdaWorldAPI]

Why this PR exists. #1294 was stacked on #1293's branch, and GitHub merged it into that branch at 19:32:16Z. #1293 had already merged to main 17 seconds earlier, so #1294's four commits (plus its merge commit) are stranded on claude/llvm-codegen-polyfill-gni3cw and are not on main:

fa9a9b09 Merge pull request #1294
04315c74 plan: mint D-CSL ids for ce64-spofc-learning-v1 + STATUS_BOARD rows
3bbbc8fd plan: ce64-spofc-learning-v1 — SPOFC into CE64, blocked legs named
951e7f67 cognitive-shader-driver: pin the emitted Pearl mask at the call site
7d8a8961 cognitive-shader-driver: emit CE64 from the SPOFC record

This PR carries exactly those commits to main. There are no new changes. The content, tests and rationale are the ones reviewed in #1294; its description applies unchanged.

What lands

Emitted CausalEdge64 (CE64) is packed from the SPOFC record:

• S and O are palette256 indices: the row's stored archetype and the partner's.
• Truth is the SPOFC evidence (a count of supporting relations), not the resonance repeated as confidence.
• The Pearl mask is now the inverse of edge_to_layer_mask. The raw predicates & 7 had turned SUPPORTS into CONTRADICTS.
• Topology bits 59–60: Direct for a P64 target, Unknown for a content-similarity partner.

Also included:

• tests/ce64_recount_probe.rs: characterises that learn() recounts a repeated observation, and that emission plasticity is ALL_FROZEN.
• Plan .claude/plans/ce64-spofc-learning-v1.md, with deliverables D-CSL-A…F and their STATUS_BOARD rows. The blocked legs name their missing dependencies.

Checks

All 10 checks were green on 04315c74 when #1294 merged. The only commit since is the merge commit fa9a9b09. CI re-runs here against main.

🤖 Generated with Claude Code

https://claude.ai/code/session_019HnekoM1EidTwQLS3oFVFm

---

Generated by Claude Code

[coderabbitai]

Navigate logical layers of code changes, visualize relationships, and explore their blast radius.

Note

Currently processing new changes in this PR. This may take a few minutes, please wait...

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Essentials

Run ID: 3ea762d9-4bff-4390-b2de-48e2e76f2727

📥 Commits

Reviewing files that changed from the base of the PR and between a1c9cae and fa9a9b0.

📒 Files selected for processing (6)

• .claude/board/INTEGRATION_PLANS.md
• .claude/board/STATUS_BOARD.md
• .claude/board/SUPERSESSION-INDEX.md
• .claude/plans/ce64-spofc-learning-v1.md
• crates/cognitive-shader-driver/src/driver.rs
• crates/cognitive-shader-driver/tests/ce64_recount_probe.rs

 ______________________________________________________________________________________________________________________
< You can't write perfect software. Software can't be perfect. Protect your code and users from the inevitable errors. >
 ----------------------------------------------------------------------------------------------------------------------
  \
   \   (\__/)
       (•ㅅ•)
       / 　 づ

✨ Finishing Touches

📝 Generate docstrings

• Commit to this branch
• Create a new PR

---

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

[cursor]

Bugbot couldn't run - usage limit reached

Bugbot is counted against Cursor usage for this user or team, and this run hit a usage or spend limit.

A user or team admin can review and increase usage limits in the Cursor dashboard.

(requestId: serverGenReqId_74995a8b-c04d-4a58-b4e7-1788f728f496)

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: fa9a9b09a2

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".

[chatgpt-codex-connector]

Keep evidence scoped to the selected object

When one row has supports to different objects or through different predicate planes, CandidateTable keeps only the highest-resonance object but unions predicates and counts support across every relationship. Packing that aggregate here therefore attributes unrelated predicate bits and inflated confidence to the single chosen S/O pair; for example, a strong CAUSES hit to target A plus a weaker CONTRADICTS hit to target B emits target A as both CAUSES and CONTRADICTS, and edges_to_layered_rows materializes the fabricated contradiction. Aggregate evidence per object, or emit separate edges, before encoding it as a binary CE64 relation.

Useful? React with 👍 / 👎.