[AdaWorldAPI]

Lab only. This is a read-only source census plus isolated falsifiers under .claude/probes/2026-11-03-d-xlog-meta-wire/probe/, which has its own workspace. Two binaries:

binary	checks
d-xlog-meta-wire-probe (P0)	18/18
v2_compose (P0b)	25/25

There is no production change. dt64, r2il, mask-risc, Quack and ndarray are used unchanged. GuardNotExecutable is untouched, and nothing adds a Gather admission, a new relation, a scope store, a scheduler or an evaluator.

P0: can M1's result reach a separately registered M2?

No, not as two methods. The first break is Receipt → M2.

• binding::run has no input for another method's result.
• A field the same run has not written is read from a resident value lane, or refused with FieldUnbound / LaneUnusable.
• Masks never cross a run:
  • the receipt keeps only the field id;
  • discard_pending turns the slot into StaleSlot;
  • KEEP is not executable.

Other refusals on the path:

• Scope { id } is an id compared for equality; it carries no population.
• A WHEN over M1's field gives GuardNotExecutable.
• VIA (the cross-row hop) gives NotExecutable and is gated by EXO-140.

WIRING-MAP.md has the full transition table.

P0b: two MethodRefs resolved at cold preparation into one program (COMPOSE-RULING.md)

No existing mechanism. Neither D-DT64-6 nor OGAR composes method bodies:

• composition today runs through a written field (DO → field → ON, KausalSpec::Depends) or stays inside one method;
• ogar-loco::Program was rejected as the method body.

The missing edge. No guard leaf can name another method's predicate as (MethodRef incl. sig_rev, field). SetExpr::Sym leaves its meaning to the consumer, so GuardWitness cannot tell Sym(1)=M1 (a<5) from Sym(1)=M1B (a<7).

Lab composer. It emulates the proposed ruling using existing contracts only: MethodBinding as the IR, a consumer-defined Sym, and register_derived + route.

requirement	result
one native program	yes
matches the oracle	177
M1's predicate disabled	435
wrong permit, class, sig rev, schema rev, scope, route, frame	refused
reference cycle	refused
predicate computed from a fold	refused
Kleene WHEN	refused
register_derived + re-route	succeeds
narrowing vs Scope	shown to be in-program predicate narrowing, not a DT64 population Scope
real DR-coordinate virtual relation	not done; separately gated

The proposed D-DT64-6 ruling is in COMPOSE-RULING.md §4. It is PROPOSED / NOT IN FORCE.

Corrected chain (operator, same day): Global WHERE → DR view → MethodRef → Fold

This is a source read only; nothing new was run.

step	status
Global WHERE (Quack Keep, 8 KiB)	production
DR addressing	production
projection and admission (dr-math-binding-probe::where_dr3::LocalSelection)	lab only, not lifted into dt64

First break (P1). An admitted selection is bound to one method: register.method != self.method → SchemaChanged (where_dr3.rs:244), pinned by an existing test. Two methods can read one WHERE only by projecting it twice. Production dt64 has only Scope { id }.

Run

cd .claude/probes/2026-11-03-d-xlog-meta-wire/probe
CARGO_PROFILE_DEV_DEBUG=0 cargo +1.99.0 run --release -q --bin d-xlog-meta-wire-probe
CARGO_PROFILE_DEV_DEBUG=0 cargo +1.99.0 run --release -q --bin v2_compose

Sources:

• lance-graph fb5e2d35b10fc4940f5c78285bae246f90c7faad
• OGAR main (checkout 5b82ae0)
• ndarray b3d7919662a3

Owner decisions

• D-DT64-6: the proposed ruling, i.e. reference-not-result plus cold inlining, including how the composite's permission is derived.
• D-DT64-6 / EXO-46: lift the WHERE-view binding (LocalSelection + Criteria) into dt64 and tie it to MethodRef, and decide whether one admitted selection may serve several MethodRefs.
• EXO-140: whether VIA / EqU32Via may be executable.
• v3 write path: a producer of change events for methods_on.

Related: #1910, #1911, #1913, #1936–#1941, EXO-46, EXO-73, EXO-108, EXO-140. The MagicSetReport and polarity work belongs to D-XLOG-WIRE (#1936) and is not claimed here.

🤖 Generated with Claude Code

https://claude.ai/code/session_01XAqKMNZqKXSaeb9g2ypywF

[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: 0d72d9ee-d435-4b4f-8776-31d867458c45

• 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

CodeRabbit Verify

• Run CodeRabbit Verify to test the changes in this pull request.

[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 2026-11-03T17:48:00.249434Z	fa7844e	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.