[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 #2078. The #2071 copy plumbing works. Run 39898485708 simply had nothing to copy.

• Why the copy was empty: both legs were stopped as superseded. The only summary line, the watcher's - Superseded: …, goes to $GITHUB_STEP_SUMMARY alone. That was a deliberate choice, so the test step never sees JOB_SUMMARY_COPY. Containment summary and Experimental summary wrote nothing on that clean path.
• Not affected: Unit Tests and Python Compatibility. The ci: report Experimental Tests and Python Compatibility to Discussions, with step summaries (#2030) #2071 PR run uploaded job-summary-test-4 and job-summary-engine-tests-*.

The fix:

• superseded.py watch gains --summary-copy PATH. Its two summary lines go to the job summary and then to that file.
• Run unit tests passes --summary-copy "$RUNNER_TEMP/job-summary.md" to the watcher only, so the tests still never see the variable.
• The comment in experimental.yml, the sentence in workflows.rst and the _actions docstring are updated.

Tests: test_superseded_summary_copy_2078 runs the real Run unit tests and Name the job summary scripts. It then reads the files at the upload step's own path:, and checks that the tests never see JOB_SUMMARY_COPY. On main, 6 of its 9 tests fail. The superseded, reporter, KTLO-guard, matrix-tier and timing tests pass.

Overlap: #2079 moves superseded.py's summary writes onto _actions.summary, so whichever merges second needs a rebase. This merges cleanly with #2076.

[JarryShaw]

GOOD TO GO at 0e31e048d (Sonnet cross-review).

• Diagnosis: confirmed from job 120236521471. The superseded notice was the only line worth a summary. Containment summary and Experimental summary wrote nothing, so Name the job summary (gated on a non-empty job-summary.md) skipped, and the upload found no files.
• The tests' env: --summary-copy goes only on the watcher's argv, before --, and watch copies its own env, so pytest never sees JOB_SUMMARY_COPY. ci: report Experimental Tests and Python Compatibility to Discussions, with step summaries (#2030) #2071's guard still passes.
• Write failures: an unwritable copy is skipped, and the watcher still stops the tests and exits 0.
• Merge order (not blocking): this conflicts with refactor(util): move the remaining Actions output onto the helper (#2047) #2079 in superseded.py. If refactor(util): move the remaining Actions output onto the helper (#2047) #2079 lands first, this PR slims down to the env line, the flag and the workflow line, since _actions.summary already tees to the copy. I'll rebase whichever goes second.

[github-actions]

Coverage: 91.54% (unit tier, Python 3.15, 53bcc8062, 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 0e31e048d: the review is GOOD TO GO and CI is complete (56 passed, 0 failed). The ordering leg ran 2050 tests with 0 failures. The 3.8–3.10 KTLO legs show 0 failing tests, and the PR is based on current main (61bb4ba66). It conflicts with #2079 in superseded.py, so whichever merges second needs a rebase. If #2079 goes first, this one slims down when I rebase it.

[JarryShaw]

Rebased onto main (26f1a72ef) and slimmed: 0e31e048d → 53bcc8062. #2079 moved superseded.py's summary writes onto _actions.summary, which already copies to JOB_SUMMARY_COPY, so this PR's own summarize() helper is gone.

• What's left in superseded.py: the --summary-copy option, and one line in watch() that adds JOB_SUMMARY_COPY to the env mapping the watcher hands to _actions.summary. The tests' environment never gets the path.
• Behaviour is the same as the reviewed version: both superseded summary lines are copied, and an unwritable copy is skipped without failing the stop.
• Tests: test_superseded_summary_copy_2078 needed no change. It and the superseded, helper, KTLO-guard and reporter tests pass.
• Also found: main was red after ci: stop warning about missing JUnit on legs stopped as superseded (#2073) #2076 and refactor(util): move the remaining Actions output onto the helper (#2047) #2079 both landed. refactor(util): move the remaining Actions output onto the helper (#2047) #2079's byte pin expected an empty $GITHUB_OUTPUT after a stop, but ci: stop warning about missing JUnit on legs stopped as superseded (#2073) #2076 writes stopped=true there. I fixed that test-only in 26f1a72.

The diff changed, so a delta review is running.

[JarryShaw]

GOOD TO GO at 53bcc8062 (Sonnet delta review).

• No leak into the tests: the new line builds a copy of env, and os.environ is untouched. After it, env reaches only report(), stopped() and the not-stopped _actions.summary call. The tests' Popen passes no env=, so JOB_SUMMARY_COPY never reaches them.
• Both summary lines are copied: the stopped line and the "not stopped" line both pass environ=env to _actions.summary. An unwritable copy is skipped without raising.
• Diff and tests: the diff covers only the expected six files. The superseded, helper, KTLO-guard, reporter and shellcheck tests pass.

[JarryShaw]

Ready to merge at 53bcc8062. The Sonnet delta review is GOOD TO GO and every blocking check passed (Required checks passed SUCCESS). The 3.8–3.10 KTLO legs carry no warnings. The ordering leg ran 2079 tests with 0 failures. The head is on current main (26f1a72ef).