ci(cron): say Saturday where the cron is Saturday-only (#964)

[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 — N/A, comment text only, no behaviour to cover
• 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 — workflows or build tooling

---

Description of your pull request and other information

Closes #964.

Both scheduled crawlers annotated 0 10 * * 6 as "everyday at 10am". The
day-of-week field is 6, so it fires Saturdays at 10:00 UTC — the schedule is
right, the comment was not. Fixed in cron-vendor.yml:5 and cron-conda.yml:8.

The repo's prose already had it right, which is what made the comments
misleading rather than merely terse: docs/source/contributing/workflows.rst:108
labels that node schedule (Saturday), releasing.rst:32 calls the crawl
weekly, and tests/_dependency_gates.py:457 says "the three Saturday schedules".

Of the four other scheduled workflows, three carry no cron comment —
codeql-analysis.yml:10, deploy-pages.yml:16, python-compatibility.yml:9.
The fourth, lint.yml:138-139, already carries a correct one. So nothing else
repeats the defect.

Comment text only. yaml.safe_load on each file before and after returns
equal structures (STRUCTURES EQUAL: True for both), so no trigger, job or step
changes. Found by the cross-review on #963; the cron-conda.yml copy turned up
on checking whether cron-vendor.yml was the only one.

[JarryShaw]

GOOD TO GO at 994e78665 — sonnet cross-review, first round. The code is
correct and complete; it disputed my own description of it, and it was right.

I claimed the four other scheduled workflows "carry no comment at all".
lint.yml:138-139 does carry one, and a correct one — it names Saturday and the
sibling schedules. Three of four carry none: codeql-analysis.yml:10,
deploy-pages.yml:16, python-compatibility.yml:9. My grep matched only the
cron: line, so a comment above it was invisible to it. Corrected in this PR's
body and in #964's. No code change — the defect really is confined to the two
files fixed here.

Re-derived independently on my side: yaml.safe_load before and after returns
equal structures for both files; no workflow sets a timezone: override, so the
Actions default of UTC holds; and tests/_dependency_gates.py:457 already called
these "the three Saturday schedules", which is further evidence the comment —
not the schedule — was the wrong half.

One fact worth recording rather than acting on: 0 10 * * 6 fires on the hour,
which GitHub documents as a high-load window where schedule can be delayed. The
comment describes the cron rule, not a delivery guarantee, so it does not
overclaim.