The changelog for an empty version will now be:

```
No user facing changes.
```

And this will appear in the final changelog when there is an actual release.
The benefits are that users will see regular release cycles and know
how old versions are even if there's no changes for a particular version

If we find that we are going months without any user facing changes, but
we have non-visible changes, then we can rethink this strategy.
But I think this is nicer than having empty sections for a version.

This can be removed when 2.5.6 is released.

This change ensures that the the script can handle
commits with no committer in them. This will happen
for some commits that are auto-generated during
PRs.

This was causing the workflow to run and create tags after merging
main into a branch.

Also, make the baseline count message less awkward sounding.

Uses two pre-existing packages to run some simple queries on a javascript
database.

gh will complain if there is an empty `""`, which is the case if
`$DRAFT` is unset.

We need to tag and mergeback if the tag does _not_ already exist.

Minor change.

Ensures that the runner version is bumped along with the action version.

Don't run the cron workflow on forks, since they lack the necessary secrets.

This commit ensures that the changelog is updated before a release with
the correct date and version.

Also, after a release, a mergeback PR is created to ensure that the
changelog update and version bump is available in main.

Adds an empty changelog file and a reminder to update it when opening
pull requests.

Also, adds a 1.0.0 version number in the package.json, which is what
we _could_ use for version numbering.

…d for runs that are not from pull requests.

Create a prerequisite job that runs the init step twice, with `tools: null` and `tools: latest`.
Use the outputs of these steps to compare the two CodeQL versions.
Pass the list of distinct tool versions for the analysis job to matrix over.
This lets us test the analysis against both versions, while avoiding duplication
when they are actually the same version.

Create a prerequisite job that runs the init step twice, with `tools: null` and `tools: latest`.
Use the outputs of these steps to compare the two CodeQL versions.
Pass the list of distinct tool versions for the integration tests to use in their matrix strategy.
This avoids redundant test jobs when the default and latest bundles are actually the same version of CodeQL.

`~` is accepted by JSON but not by the Actions context language, so we use `null` to indicate the default version.

…s: latest`

Always test against both the default and latest CodeQL bundle.

This improves test coverage shortly after a CodeQL bundle release, where the latest bundle
may not yet be built into the Actions VM image as the default bundle.

It also saves a manual step during bundle release testing,
since we no longer need to temporarily change the PR checks to `tools: latest`.

There is some redundancy when the latest bundle is the same as the default bundle on the VM image,
but this can be considered a test for the `tools: latest` configuration.