D-GSO-6: deterministic, versioned recipe selector (P6)

[AdaWorldAPI]

What

P6 of .claude/plans/2026-10-06-global-sudoku-replayable-orchestration-v1.md: recipe selection as a pure, versioned function of declared epistemic state (§11 loop, §13 replay contract, §18 P6).

crates/cognitive-shader-driver/examples/recipe_selector_probe.rs, no library code changes.

• SelectorPolicy::select(state) is a const fn: no clock, counter, global or randomness.
• The declared state is four facts an earlier step already measured: observations pending, frontier bounded, local disagreement, new encounter.
• V1 follows the §11 order (fold → bound → local → revise) and returns no recipe only when nothing is open, so a hole is never forced into a guess.
• V2 swaps two steps. It exists to show a policy change is a new version: V1 and V2 differ on exactly 2 of the 16 states.
• A recorded Selection (policy + state) replays to the same recipe from the record alone, even after the current default changes.

Scope limits

• Recipe ordinals and the V1 order are a scaffold; ProbeRecipe is not a shipped recipes::Recipe ID (plan §12).
• The policy version covers selection only. Plan §13 requires it to cover every executable operation a recipe can select; that semantics digest is not done here.
• The replay test's state transitions are a stand-in: each recipe clears the condition it answers.

Tests (5) and disable runs (5, all red)

• V1 matches a written-out table oracle on all 16 states — D2: swap V1's order; D5: rest with an encounter still open
• selection is the same across repeated and reversed call order — D1: an impure selector with a hidden call counter
• a recorded selection replays under its own policy, including on a state where V1 and V2 differ — D3: make V2 identical to V1
• every cycle rests in exactly one step per open condition and replays identically — D4: a recipe that does not clear its condition
• only the fully settled state rests

Board

STATUS_BOARD: D-GSO-6 In PR; D-GSO-5 Shipped (#1355).

🤖 Generated with Claude Code

https://claude.ai/code/session_01DgxrCafsJuAahpBs7oC14R

---

Generated by Claude Code

[coderabbitai]

Warning

Review limit reached

Your organization has reached its usage spending cap. Adjust your spending cap in the billing tab.

Next included review available in 33 minutes.

Check out review usage here.

View limit details

Limit details: You’ve used the included review currently available. Your 67 included PR review attempts over the past 7 days set your current allowance at 1 review per hour.

Learn how review limits work.

Review configuration:

⚙️ Run configuration

• Configuration used: Organization UI
• Review profile: CHILL
• Plan: Essentials
• Run ID: 620b4fa9-5d72-42ff-923f-a131d381c44a

📥 Commits

Reviewing files that changed from the base of the PR and between fdb08b2 and e4fb835.

📒 Files selected for processing (4)

• .claude/board/STATUS_BOARD.md
• crates/cognitive-shader-driver/Cargo.toml
• crates/cognitive-shader-driver/examples/README.md
• crates/cognitive-shader-driver/examples/recipe_selector_probe.rs

• Autopilot · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Warning

Billing warning: we have not been able to collect payment for this subscription for more than 72 hours. Please update the payment method or pay any pending invoices in Billing to avoid service interruption.

---

_{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: eaa0f504-4f37-42a3-8ff4-80498cd31633)

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-10-06T11:31:57.354240Z	8a01fa0	Draft marked ready

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" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.