Add support for --help

[Yurii201811]

Closes #4119.

Summary

• route the exact task --help form through the existing help command
• verify that --help and help produce the same stdout and stderr
• assert that the help output contains the expected usage text
• document --help in the task man page

Verification

• uvx pre-commit run --all-files
• cmake -S . -B build -DCMAKE_BUILD_TYPE=Debug
• cmake --build build --target test_runner task_executable --parallel 8
• ctest --test-dir build -R '^commands\\.test\\.py$' --output-on-failure (1/1 passed)
• ctest --test-dir build --parallel 8 --output-on-failure (178/178 passed)
• cargo fmt --all -- --check
• cargo check --locked
• cargo clippy --all-features --locked -- -D warnings
• cargo test --locked (10/10 passed)
• .github/workflows/metadata-check.sh
• source package generated, extracted into a clean temporary directory, configured, and rebuilt through task_executable
• git diff --check origin/develop

[Yurii201811]

CI note: formatting, pre-commit, Clippy, Cargo metadata, Cargo tests, and tarball packaging all passed. The coverage job also completed its build, tests, and report generation; only the final Coveralls upload failed with a transient 504 Gateway Timeout. GitHub does not allow fork contributors to rerun that job, so there is no contributor-actionable failure in this patch.

[swordskode]

@Yurii201811 You're just submitting AI code as your own?

[Yurii201811]

Yes, I used Codex as an implementation assistant. I selected issue #4119, supervised and reviewed the change, and I’m responsible for the PR and any follow-up. The repository’s contribution guide does not appear to prohibit or require labeling AI-assisted contributions. CI ran all 178 tests successfully; only the external Coveralls upload failed with a 504.

If you have a specific concern about the implementation, test coverage, or project fit, please point it out and I’ll address it.

[swordskode]

@Yurii201811 I would suggest you do the work yourself.

If you do not have the time or the skill to do the work, don't do it, and leave it for those in the community that do.

[Yurii201811]

Thanks for clarifying. I don’t share the view that responsible AI-assisted development means a contributor lacks skill or hasn’t done the work. I supervise the process, review the changes, and take responsibility for them. Since no concrete technical concern has been raised, I’ll leave the PR for the maintainers to evaluate on its merits and won’t continue debating the tooling.

[swordskode]

@Yurii201811

But, you haven't done anything...

These projects are maintained and built by a community of engineers that have the skills and time to do so.

Unfortunately there are people, like yourself, that look to use AI tooling to make "contributions", to open-source projects, in order to bolster their profiles.

You undermine your development as an engineer, undermine the community as a whole by submitting "contributions" like this.

[Yurii201811]

@swordskode You do not know my professional background, motivations, or the work I did to understand, review, and validate this change, so please stop making personal allegations. Used responsibly, AI empowers professionals; it does not replace judgment, skill, or accountability. I hope you remain open to learning how these tools can be leveraged responsibly as well. If you have a specific technical or project-policy concern, state it clearly; otherwise, I’ll leave this to the maintainers and won’t engage further.

[djmitche]

Taskwarrior and GothenburgBitFactory do not have a written AI policy, but the existing open-source standards apply: contributed code is assumed to not carry contradictory licensing restrictions, and in the review process we expect the contributor to understand and modify the patch as necessary and not expect maintainers to do extra work to make their patch work.

None of that is the case here: this is a small patch that does the necessary thing and has all of the responsible bits I might expect (docs, tests). Sure, the PR summary is a bit chatty, but that's not hurting anything.

We do have a very brief code of conduct:

Be respectful of all contributors, all contributions, and everyone’s time, in all project-related forums. No exceptions.

The comments here are not in the spirit of that code: regardless of our feelings on AI, comments directly about other contributors are disrespectful.

Finally, I'll note that there are, in fact, only a few active developers in this community, so aside from adherence to a code of conduct, a respectful and supportive environment is critical to broadening that group of developers. @Yurii201811, your work is welcome here, whatever tools you use to do it.

[djmitche]

This is quite a succinct way to accomplish this!