[AdaWorldAPI]

Two commits, one per concern. Review them separately.

1. dca11fd: #1650 closeout (board files only)

• STATUS_BOARD: D-CMR-2 marked Shipped (D-CMR-2: lower the contract set algebra onto mask-risc (runs as extents, three levels); D-CMR-5 re-scoped #1650, f031089).
• PR_ARC_INVENTORY: new entry for D-CMR-2: lower the contract set algebra onto mask-risc (runs as extents, three levels); D-CMR-5 re-scoped #1650.
• LATEST_STATE: new "main after D-CMR-2: lower the contract set algebra onto mask-risc (runs as extents, three levels); D-CMR-5 re-scoped #1650" section.
• The D-CMR-2 entry is marked merged. Codex's review of D-CMR-2: lower the contract set algebra onto mask-risc (runs as extents, three levels); D-CMR-5 re-scoped #1650 found nothing.

2. 2e71382: D-CMR-6, the guarded rule catalogue and fixpoint driver

New zero-dependency module lance_graph_contract::set_rules.

• Measures and semantics. A measure is Count, Sum, CountRows, Add, Sub, Const or Opaque. The semantics are explicit: the row count (universe), the arithmetic (exact, wrapping, checked, IEEE), two- or three-valued sets, plain or _sym sums, and distinct rows or a bag.

• Rules. There are six, stored as data in CATALOGUE:

  • set normalization;
  • Count(¬A) → N − Count(A);
  • Sum(A) − Sum(A∧B) → Sum(A∧¬B), and the same for Count;
  • (a+b) − a → b;
  • range intersection;
  • range count.

• Guards. Each rule refuses when its conditions fail:

  • NULL (three-valued sets);
  • no fixed row count;
  • checked or IEEE arithmetic;
  • _sym sums;
  • an impure operand;
  • ranges on different row orders;
  • a range bound past the row count;
  • a bag.

  A refused rewrite keeps the original expression and is recorded in the trace.

• Driver. It applies only rewrites that strictly shrink the expression, and stops when no rule fires or the step budget runs out.

• Tests (17).

  • A random test runs 600 random measures over random tables against exact i128 and wrapping i64 reference evaluators, with random subsets of the rules enabled. It also checks that most inputs are actually rewritten.
  • Experiment A: under checked arithmetic the difference rule refuses. An eager i32 accumulator is the counterexample: the original overflows, while the rewritten sum gives 5.
  • Experiment D: intersecting three ranges enables the range count.
  • Experiment E: every invalid case above refuses. Each NULL case is also shown giving a wrong answer if the rewrite is forced.

• Disable runs. I removed each guard in turn, and each removal fails exactly its named test.

• Gates. All 1,539 contract tests pass. The 26 relalg tests pass unchanged. clippy -D warnings and fmt are clean.

• Untouched: set_algebra.rs, set_program.rs (U2), crepe (U3), relalg and mask-risc.

Board, updated in the same commit: STATUS sets D-CMR-6 to In PR; new entry entries/2026-10-21-d-cmr-6-rule-catalogue.md; the LATEST_STATE inventory delta; the plan's gap ledger.

Coordination: claimed on EXO-50 as session_013z2e1YXkm6o6HNrDn5BPWA/016; ownership confirmed in gpt-ada-01/004 and /007.

🤖 Generated with Claude Code

https://claude.ai/code/session_013z2e1YXkm6o6HNrDn5BPWA

[coderabbitai]

Important

• 🔍 Trigger review

This repository does not receive automatic reviews because it has fewer than 10 stars.

⚙️ Run configuration

• Configuration used: Organization UI
• Review profile: CHILL
• Plan: Advanced
• Run ID: ac8db9c5-771d-4039-ba36-67f6652ef6e9

• Autofix · 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.

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.}

[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 2026-10-21T07:28:36.197241Z	2e71382	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.

[coderabbitai]

Add Carrot credits or activate Agent usage billing to use Autofix

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 2e71382a67

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".