Fix layering gate false edges: reference-derived producer misattributes prelude bare names (main-red)

[gunbai-bot]

Auto-opened by session-dashboard for session wise-pike-555.
Pushing to session/wise-pike-555 advances this PR.

Worker attestation

Before flipping this PR to ready for review, confirm each item:

• Title describes the change (not the session id or branch).
• PR body summarises what and why (replace the TODO below).
• Tests run: name the command (e.g. npm test, cargo test) and the result.
• If this closes a work item, the body contains a Closes #N directive.
• No commits on this branch are surprises (no fork/cherry-pick I did not make).
• No secrets / credentials / large binaries staged.

Summary

TODO: replace this paragraph with one or two sentences naming the change and its motivation. Reviewers read this first.

Test plan

• TODO: list the commands that ran (or "no tests changed; relied on CI") and the outcome.

[gunbai-bot]

review 41546 (RecordLit / skip_bare_name propagation) — verified, no change needed.

The hypothetical under-collection (Foo { bar: baz } missing baz) does not occur with our AST shape. parse_field_init / make_field_init_node stores the label on the field-init node's name and the value as children[0] (field_init_node_value), not as a bare name on the property node itself.

Traversal for ExprRecordLit visits field-init children with skip_bare_name=true (so the label bar is skipped), but collect_node_refs_inner always recurses into node.children with skip_bare_name=false at the bottom of the function — so the value subtree (baz, including ExprCall / qualified chains) is still collected. The early return on ExprRecordLit only prevents the record node's own fallback child walk from double-visiting; it does not block value collection inside field-init subtrees.

Agree this is safe-for-loader under-collection at worst; in practice the shape does not arise here.

review 41547 — no findings; APPROVE accepted as-is.

— sent from wise-pike-555