OpenSpec Workbench: running and supervising agents on OpenSpec changes

[VeryComplexAndLongName]

I have been building a tool on top of OpenSpec for my own workflow, and it is
listed in the Community Showcase under its old name and old description
("OpenSpec UI", a dashboard for browsing changes). It has since grown into
something different, so I would like to correct the entry. Before I open a
pull request for that, I wanted to ask whether the wording below is right for
the page.

What it is: OpenSpec Workbench runs a coding agent CLI (Claude, Copilot,
Codex, Gemini or a local model) on an OpenSpec change, shows what the run is
doing while it does it, and lets a person stop it where its work is sound. It
is a standalone local web application and a VS Code extension over the same
core. It reads OpenSpec's files and runs your own agent CLIs; it needs no
account and no paid service, and it never handles an API key.

What it does not do: Codex and Gemini support has never been run against the
real binaries, only against mocked peers. The repository and packages are
still named OpenSpec-UI.

Proposed line for docs/community.md:

• OpenSpec Workbench: Runs and supervises coding agents on OpenSpec changes, from a local web app and a VS Code extension. Free; runs on your machine with your own agent CLIs.

#1934

Disclosure, as CONTRIBUTING asks: I develop this with Claude Code, and the
text of this post was drafted with its help and checked by me.

Would you like the entry changed in any other way?