[AdaWorldAPI]

This PR has two commits. One adds the G6 lab work: CanCook executed from its DT64 MethodRef. The other records the merged Foundry parity work on the board. No DR or DT64 contract changes.

G6: a registered method, lowered once per binding (fcee68c)

World::new now registers one MethodBinding for each (item, quantity) pair, with quantity ∈ {1, 3}. Each method's WHEN guard says "this store can make q of item m" as a SetExpr over the class's field positions: mask fields as Sym(field) and value fields as residuals. The method's reads are exactly the fields its guard names. DT64 guards take no parameters, so the quantity is a constant of the guard and part of the method id.

PreparedMethod::prepare does the following once per binding:

• admits the MethodRef (BindStore::admit);
• binds and borrows the reads plus storeId;
• renames field positions to the bound planes, and conjoins the reader scope. The reader scope is the IAM stand-in and is never part of the method;
• lowers the guard once through relalg and keeps the returned mask-RISC program.

run only executes that program. The lab writes no lowering of its own.

relalg: adds rel_from_set, a public wrapper over the SetExpr → Rel conversion that run_program already applied (from_set stays crate-private), plus a round-trip test.

Contract dependency: OGAR pulls lance-graph-contract from git, so the lab now patches it to the path crate, as relalg does. The lab's graph holds one contract and one ndarray.

Refusing a multi-fold lowering: relalg keeps only the last fold's program, so prepare refuses a guard that lowers into more than one fold. That cannot happen today; a test pins that every guard shape lowers to one fold.

Evidence

There are 8 new tests, 32 in the lab overall:

• all 16 methods match the oracle;
• row order does not change the answer;
• the reader scope is load-bearing;
• admit refuses a forged guard, another signature revision, and another store's reference;
• a method whose reads miss a guard field is refused with NotBound;
• repeated runs of one prepared method give the same answer;
• every guard shape lowers to one fold.

Seven disable runs each turned the suite red and were restored:

• drop the staff conjunct;
• drop the known conjuncts;
• make the stock threshold off by one;
• drop the reader scope;
• skip the identity mapping;
• lower All instead of the guard;
• treat an unbound field as All.

Measured

cargo run --release --manifest-path crates/foundry-parity-lab/Cargo.toml --bin bench. One VM, single thread, median of 41 runs, four runs:

Stage	Time
method_ref + admit	~5 µs
prepare (admit, bind, borrow, one lowering)	14.9–15.2 µs
CanCook question, cached program	359–400 ns
arm B, hand-written program, same question	474 ns

The first version called run_program for every question, which re-lowers each time, and took about 21 µs per question. A fifth bench run, taken while the machine was loaded, measured 600 ns and is excluded.

Open

• G6 in production: nothing yet owns a prepared program across questions.
• G5 (bound relations) waits on ADMISSION B (EXO-140).
• G1's limiting resource (P1b) waits on the min/argmin decision.

Board record (7323855)

• PR_ARC_INVENTORY entries for D-FOUNDRY-PARITY-1: evidence matrix + 64-store read-only operational lab (DT64 + mask-RISC) #1991 (637d638) and foundry-parity-lab: per-store cost via INT8 packed GEMM (D-FOUNDRY-PARITY-1, G1 partial) #1995 (15d4b4c).
• A LATEST_STATE row.
• STATUS_BOARD: D-FOUNDRY-PARITY-1 is Shipped, and D-FOUNDRY-G1 is Partial.
• An INTEGRATION_PLANS link.
• A new entry, D-FOUNDRY-PARITY-1-P1A, covering arm C and the lance-graph-java crosswalk summary, including the R11 layout-witness gap (D-DRISA-M1).

The entries and supersession indexes are current, and board files only grew. Linear: EXO-227 (mirror) and EXO-223.

---

🤖 Generated with Claude Code

https://claude.ai/code/session_01XAqKMPJjVvasjwqhjJ6c9p

[coderabbitai]

Important

• Trigger review

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

Run configuration

• Source mode: Auto
• Repository: Unavailable
• Branch: Unavailable
• File: No configuration file from this repository applied
• Commit: Unavailable
• Layers: Organization UI
• Review profile: CHILL
• Plan: Advanced
• Run ID: 131212d4-3daa-4910-a14c-24186a2c4775

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

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

[chatgpt-codex-connector]

Codex Review Summary: ✅ Completed

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

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-11-10T20:20:29.753715Z	7323855	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.