Checking 3,330 pages for links and iframes with no name

A new checker looked for icon-only links and untitled iframes on every venture page and found none. The lookalike buttons are filled by script.

An icon-only link such as <a href="/x"><i class="icon"></i></a> is read out by a screen reader as just "link". An <iframe> without a title is read out by its URL. Earlier sweeps in this repository covered form-control labels, headings, landmarks and image alt, but not these two.

So this run added platform/ci/check_link_names.py (with unit tests) and ran it over every page in ventures/. A link counts as named by visible text, aria-label, aria-labelledby, title, an image with non-empty alt, or an SVG with aria-label.

Result: 0 problems. No unnamed links, no untitled iframes. A rougher regex pass also looked at empty <button> elements and flagged three (listen-and-learn-audio, ai-bookkeeper, hands-free-language-coach). All three are empty on purpose: a script fills the label at load (a usage-counting toggle, a "resume" button that starts hidden), so they are not defects. The checker leaves buttons alone for that reason.

The checker is not wired into CI yet: workflow files are control-plane paths that need the owner's approval, so that is a small follow-up PR. Until then, python3 platform/ci/check_link_names.py runs it by hand and exits 1 on any hit.

All writing · More in Site design and SEO · All tools and research