fix: spell product names and acronyms correctly in tag titles

[abrichr]

Tag pages on the blog misspelled product names and acronyms. For example, the UiPath tag page was titled "Uipath" and the RPA tag page "Rpa". This pull request gives each affected tag a term page with the correct title, and adds a test that fails when a published post uses an acronym or brand tag that has no term page.

Fixes

Problem	Change	Regression test	Ref
Hugo built tag titles from the lowercase slug, so eight tag pages showed misspelled names: Api, Autohotkey, Gui-Automation, Openadapt-Flow, Openemr, Power-Automate, Rpa and Uipath. The wrong names appeared in each tag page's title and heading, in the tag links on posts, and in the tag RSS feeds.	Added content/tags/<slug>/_index.md for each of the eight tags, with the titles API, AutoHotkey, GUI Automation, openadapt-flow, OpenEMR, Power Automate, RPA and UiPath.	tests/test_tag_titles.py runs in the existing python3 -m unittest discover -s tests CI step. It fails if a published post uses an acronym or brand tag that has no term page with the expected title, or if a term page has no published posts, which would publish an empty tag page. It reads both YAML and TOML front matter, so drafts created with hugo new don't break it. When Hugo and the theme are installed, it also builds the site and checks each tag page's <title> and <h1>.	B238

Not fixed here

• The tag list at /tags/ still shows every tag by its lowercase slug, for example "uipath" and "rpa". The theme lists terms by slug rather than by title, so the list is consistently lowercase rather than misspelled. Showing titles there would need a theme layout override that changes how every tag looks.
• Generic hyphenated tags such as "browser-automation" and "computer-use" keep Hugo's default titles ("Browser-Automation", "Computer-Use"). They aren't misspelled names, but on a post that also has the GUI automation tag, the tag styles differ.

How it was tested

• On main, the new test fails: the source check reports the eight missing term pages, and the build check reports titles such as Uipath | OpenAdapt Blog. On this branch, all tests pass.
• Without Hugo installed, as in the CI test job, the build check is skipped and the source check still fails on main and passes here.
• A full hugo --minify build produces the same set of files as main. Only the eight tag pages, their RSS feeds, /tags/index.xml and the posts that link to those tags change, and only in tag names.
• Voice lint, substance lint, check_benchmark_claims.py and validate_newsletter_render.py pass.
• Merged on top of fix: correct post claims, replace the dead signup form, rewrite posts #59 (fix/blog-overhaul), the full test suite passes. That pull request moves several posts to draft, but each tag that has a term page still has at least one published post.

Before merging

• This pull request and fix: correct post claims, replace the dead signup form, rewrite posts #59 change different files, so they don't conflict. If fix: correct post claims, replace the dead signup form, rewrite posts #59 merges first, rerun CI on this branch, because fix: correct post claims, replace the dead signup form, rewrite posts #59 changes which posts are drafts and the orphan check depends on that.
• The draft post "The keystroke that lied to us" uses an rdp tag. Before it's published, add content/tags/rdp/_index.md with the title "RDP" and an "rdp": "RDP" entry in EXPECTED_TITLES, or the test fails.
• Optional, after fix: correct post claims, replace the dead signup form, rewrite posts #59 merges: add a step to .github/workflows/deploy.yml after hugo --minify that runs the build check against public/, so CI also checks the rendered titles. fix: correct post claims, replace the dead signup form, rewrite posts #59 changes that workflow, so this pull request leaves it alone.

🤖 Generated with Claude Code