[JarryShaw]

• Searched for similar pull requests
• Followed the coding style (make pylint, make mypy, make isort)
• make test passes, and a test case covers the change
• Added a changelog entry — N/A — centralised in docs(changelog): shared 1.5.0 changelog — long-lived, merges last (#610, #616, #617, #618, #620) #657

What is the purpose of your pull request?

• ci — changes to CI configuration files and scripts

---

Description of your pull request and other information

Closes #2077. When the watcher stops an Experimental leg as superseded, Containment summary now says tests were stopped as superseded after N s instead of tests exited 0, which read as a pass.

• How: the step gets STOPPED: ${{ steps.tests.outputs.stopped }} (from ci: stop warning about missing JUnit on legs stopped as superseded (#2073) #2076) through env:, never inside run:, and only the closing line changes.
• No new warning: the watcher's Superseded notice already explains the stop.
• Unchanged: every other warning (No memory cap, OOM kill, timeout, MemoryError, xdist worker) and every count. Job 120233865230 shows a stopped leg reports 0 worker(s) down and no warnings, so there was nothing else to suppress.
• Docs: workflows.rst gets one sentence on it.
• Tests:
  • New: test_superseded_containment_summary_2077 runs the real Containment summary script for a stopped run and a normal run. It also runs it after the real Run unit tests step, once stopped and once on a current commit. On main it gives 6 failed.
  • Updated: test_resource_monitor_1740's env pin now includes STOPPED.
  • Passing: the superseded, KTLO-guard, matrix-tier, resource-monitor, reporter, actionlint and shellcheck tests.
• Overlap: it touches only the Containment summary step, and merges cleanly with refactor(util): move the remaining Actions output onto the helper (#2047) #2079 and ci: copy the superseded watcher's summary lines into the job-summary artifact (#2078) #2080.

[JarryShaw]

GOOD TO GO at 6a82fae3a (Sonnet cross-review).

• Workflow diff: a YAML diff shows only the Containment summary step's new STOPPED env key and its run: body changed.
• The output reaches the step: Run unit tests has id: tests and continue-on-error, and Containment summary runs later in the same job under !cancelled(). So the output is readable even after a failure. When there was no stop, $STOPPED is empty and the summary still says exited ${code}.
• Real problems still warn: every warning (No memory cap, OOM kill, timeout, MemoryError, xdist worker) is raised before the new if and still uses $code. A test pins that the warnings are identical whether the run was stopped or not.
• The tests: with main's experimental.yml swapped in, the new test's stopped-run and env checks fail. Shellcheck and actionlint pass.

[github-actions]

Coverage: 91.54% (unit tier, Python 3.15, 6a82fae3a, Unit Tests run success)

Package	Statements	Missed	Branches	Partial	Cover
pcapkit (top level)	104	4	20	2	95.16%
pcapkit/const	19599	880	2618	866	92.12%
pcapkit/corekit	2565	71	812	37	96.56%
pcapkit/dumpkit	282	3	102	3	98.44%
pcapkit/foundation	2930	23	1022	25	98.79%
pcapkit/interface	138	4	52	3	96.32%
pcapkit/protocols	17414	219	4574	185	98.14%
pcapkit/toolkit	869	36	298	10	95.20%
pcapkit/utilities	473	7	136	5	98.03%
pcapkit/vendor	4432	1988	1018	167	51.87%

Per-file detail: the coverage-html artifact of this run.

[JarryShaw]

Ready to merge at 6a82fae3a. The review is GOOD TO GO, CI is complete (56 passed, 0 failed), the ordering leg ran 2058 tests with 0 failures, and the 3.8–3.10 KTLO legs show 0 failing tests. The branch is behind main by #2079, #2086 and 26f1a72. I merged it into current main (26f1a72ef) in a scratch tree, and the Containment summary, superseded, resource-monitor, shellcheck, KTLO-guard and actionlint tests all pass there.