[CI] (P7): tighten renamed CI titles

This commit is contained in:
2026-09-23 13:23:09 -04:00
parent d7b9218618
commit 877c531a0d
3 changed files with 26 additions and 26 deletions
+2 -2
View File
@@ -1,8 +1,8 @@
#!/bin/bash
# THE RETRACE-SPLIT LANE'S NEGATIVE CONTROL: a PULL library must red this split retrace.
#
# This file is the body of .github/workflows/test.yml's "Negative control - the PULL library must
# red this split retrace" step, extracted for the reason given at the top of
# This file is the body of .github/workflows/test.yml's "Negative control (pull library)" step in
# the retrace-split job, extracted for the reason given at the top of
# scripts/ci/split_negative_controls.sh: a `run:` block is unreviewable and untestable off a
# runner, and scripts/ci/control_smoke_test.sh now runs THIS file rather than a hand-made copy.
#
+8 -8
View File
@@ -1,14 +1,14 @@
#!/bin/bash
# EXIT GATE E1's NEGATIVE CONTROL and EXIT GATE E3(a)'s.
#
# This file is the body of .github/workflows/test.yml's "Negative controls - the verb barrier and
# the persistent-map push must be load-bearing" step. It lives in the repository rather than inline
# in the workflow for one reason: a workflow `run:` block cannot be executed anywhere except on a
# runner, so the logic below was unreviewable and untestable until it ran in CI - and when the
# wave-1 cross-family review claimed it was broken, confirming the claim needed a hand-made copy of
# these lines with their inputs stubbed (wave1-codex-verify.md 8). A copy is not the thing. The
# smoke test at scripts/ci/control_smoke_test.sh now runs THIS file, so the lines CI executes and
# the lines the smoke test proves are the same lines.
# This file is the body of .github/workflows/test.yml's "Negative controls (barrier, map push)"
# step, which requires the verb barrier and the persistent-map push to be load-bearing. It lives in
# the repository rather than inline in the workflow for one reason: a workflow `run:` block cannot
# be executed anywhere except on a runner, so the logic below was unreviewable and untestable until
# it ran in CI - and when the wave-1 cross-family review claimed it was broken, confirming the
# claim needed a hand-made copy of these lines with their inputs stubbed (wave1-codex-verify.md 8).
# A copy is not the thing. The smoke test at scripts/ci/control_smoke_test.sh now runs THIS file,
# so the lines CI executes and the lines the smoke test proves are the same lines.
#
# WHAT THE REVIEW FOUND (ID-46 finding 8, CONFIRMED by execution; ID-48 assigns it here).
# The previous version accepted ANY non-zero ctest exit as "the knob is load-bearing". A timeout, a