Assay

A page you have just written has no tests yet. Assay serves the page locally, opens it in a real Chromium browser through Playwright, finds every control on it, uses all of them, and reports where the page contradicts itself. There are no tests to write, no API key and no network, and the same page gives the same result every time. This skill tells you to run assay after you create or change a web page and to report the one line it prints. The skill reports the result and does not change code because of it unless you ask.

Use it when you have just created an HTML page and want to know whether its controls work, when you changed a page or the JavaScript or CSS it loads, or when the user asks you to check, QA or smoke-test a local web page. Run it once at the end, after the last edit, on each page you changed. Do not run it while you are still editing.

Know its limits. Assay will not build your project, so point it at built output such as dist/index.html; given a source tree it says so and names the command to run, and it will not run npm install because installing dependencies executes their setup scripts. It checks one page and does not crawl. It cannot check intent: a control that runs but does the wrong thing usually is not caught. A finding is a place to look, not a confirmed bug, since a control can correctly do nothing in the state it was pressed in.

Install the CLI once from PyPI. The package is named assay-ui and the command is assay. It needs Python 3.10 or newer, and on first run it downloads the Chromium build Playwright uses.

pip install assay-ui

The input is the page itself, such as todo.html or dist/index.html. A folder also works and opens the index.html inside it. The page is often a folder or two above the file you edited, so a change to a script or stylesheet still means running assay on the page that loads it. A small page takes seconds and a busy one can take a minute or more.

The output is a single line, or one line plus a short list. A clean run reads like "assay: checked todo/todo.html, 8 checks, nothing flagged." A flagged run reports the count followed by a list of places to look. Paste the line verbatim as the last thing in your reply, whether or not anything was flagged; without it the reader cannot tell a clean page from a check that never ran. If the command did not run, say so, name the page and quote what the shell printed. Never describe a check that did not happen as one that passed. When something is flagged, report it and stop; fix code only if the user then asks.

Assay uses Playwright and Chromium and installs from PyPI. It runs locally and makes no network calls of its own, serving the folder over loopback rather than opening the file directly. It is read-only toward your code and writes nothing to disk unless you pass the report option to save an HTML report. Because it drives every control it finds, a page wired to a live backend can trigger that backend, so point it at a local page or build, not a production URL. Related skills mentioned are mock-hunter, which audits a live page for mock data, hardcoded values and broken endpoints, and accesslint-diff, which diffs accessibility violations against a baseline.

Developers who generate or edit web pages quickly and want a fast, deterministic sanity check that the controls on the page work together, without writing a test suite first. It also suits teams that want a strict reporting habit: the exact line assay printed, nothing summarized, nothing invented.