source-control:babysit-prs: fetch-all-pr-comments.sh reports in_reply_to_id=None for properly-threaded inline review replies

[kyle-sexton]

Lane: babysit (safe/gate tier, /loop /source-control:babysit-prs)

Triggering example: worker processing #563 needed to confirm whether kyle-sexton's classification-table replies were threaded under the correct codex inline review comments (P1→#572 defer, P2→#573 defer) before deciding resolve-vs-leave-open. fetch-all-pr-comments.sh reported in_reply_to_id=None for these replies even though GraphQL thread membership confirmed they are properly threaded inline replies.

Observed: in_reply_to_id field from the script is unreliable for thread linkage — a consumer following the script's own contract would wrongly conclude these are top-level/unthreaded comments.
Expected: the field either reflects true reply-parent linkage, or the script's docs explicitly warn not to use it for thread-membership decisions (forcing callers to a GraphQL cross-check, which is what the worker did as a workaround this cycle).

Category: bug
Impact: any babysit worker or the orchestrator itself that trusts in_reply_to_id for "is this reply threaded under finding X" decisions can silently misclassify addressed-vs-unaddressed findings.

Work-class: C3 (bug-fix-shaped) — attended triage 2026-07-23, operator-ratified. 🤖

[kyle-sexton]

This was generated by AI during triage.

T1 — routing status: ready. Minimum viable fix has no design decision: document that in_reply_to_id is unreliable for thread-membership decisions and callers must GraphQL-cross-check (what the worker already did as a workaround). A deeper fix (making the field itself reliable) is a bonus, not required for this to ship — implementer's call whether to attempt it.

[kyle-sexton]

This was generated by AI on operator direction.

Superseded — this item was already ratified and should not have been queued.

The earlier first-drain ratification request on this issue was raised in error. The admission gate dispatches a C3 bug-fix-shaped item when it carries either its own recorded ratification or a completed first drain — and this issue already carried recorded ratification in its body from attended triage on 2026-07-23. The drain cycle applied the blanket first-drain rule without first reading bodies for an existing ratification, so it queued an item that was already admitted.

Restored to the autonomous frontier: agent-ready re-applied, needs-human removed. No human action is required, and this issue no longer appears in the attended queue's ratify view.