docs: list unreviewed master commits as a mistake

A CHANGELOG polished in a back-and-forth with an AI assistant and
committed straight to master skips the usual QA, and leaves master
with a string of commits carrying prompts nobody needs.

Prompt: Do a QA on my sentence, commit it directly to master, rebase when #768 is merged, and I'll push it manually [= the sentence added to the list of release mistakes; #768 is the towncrier PR]

Prompts (comments made inline in the document while reviewing it):
> (on the review of #768) Please piggyback my latest commit to the RELEASE-HOWTO.md into the same PR

Assisted-By: Claude Opus 5.5 (Claude Code)
Reviewed-by: Tobias Brox <tobias@redpill-linpro.com>

Author: tobixen

docs/design/RELEASE-HOWTO.md

@@ -62,3 +62,4 @@ This is most likely not complete, but should explain some of the "silly" steps a
 * Pushing out junk files in the pypi-release (i.e. .pyc-files, log files, temp files, `tests/conf_private.py`, `tests/caldav_test_servers.yaml`, an entire `venv/`, etc).  `tox -e package` now catches this - see the build step above
 * Not adding the release to the "github releases" (I don't care much about this feature, but apparently some people check there to find the latest release version)
 * Not committing after running `towncrier build` (or, up to and including 3.5.0, after doing a s/unreleased/$VERSION/ on the CHANGELOG)
+* "Playing ball" with Claude on the CHANGELOG and committing the changes directly to master before a release, not going through the regular QA - and suddenly the master branch has 6-7 commits with various irrelevant prompts attached.