docs(qa): teach the QA skills to test a configurable surface - #2584
Merged
Conversation
|
Important Review available on request
Reviews should be triggered manually for repositories with fewer than 10 stars. Select Trigger review above or comment ⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Plus Run ID: Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
ktursunov
changed the title
ktursunov
force-pushed
the
codex/report-testing-skills
branch
from
August 17, 2026 06:37
d0f649a to
c424bc6
Compare
ktursunov
changed the title
added 4 commits
August 17, 2026 15:27
Signed-off-by: Konstantin Tursunov <Konstantin.Tursunov@constructor.tech>
Signed-off-by: Konstantin Tursunov <Konstantin.Tursunov@constructor.tech>
The guidance named a "strict hook" as what a deep-link renders into. That
phrase already means something else here — a context hook that throws outside
its provider — while URL state is handled by `validatePortalSearch`, which
drops what it cannot parse, and `assertDateRange`, which throws into
`AppErrorBoundary`. A reader grepping the old phrase landed in the wrong file.
Naming all three makes the two paths a malformed link can take checkable.
`drive-ui` claimed the frontend nginx serves `/api`. It serves neither `/api`
nor `/auth` — both fall through `try_files $uri $uri/ /index.html`, which is
the actual reason a browser on that port loops. Its exploration section also
argued against the skill's own opening line ("neither is about clicking") and
carried none of the moves the skill owns; the opening now names three concerns
and the section starts from `snapshot`, `requests` and the code behind the
controls.
Suite placement went vague where the repo has tags: "frontend tests" and
"unit/component suites" are `fe-unit` and `fe-component`. `stand-ui-test` adds
that the period boundary it points at is still unclaimed — the api suite clamps
at 399 days rather than asserting 400 — and that the three shipped download
journeys stop at suffix and size by history, so their assertions are not the
model to copy.
`scope-feature-tests` gets its quoted trigger phrasings back. Compressing the
description had removed every verbatim query a user actually types, and the
description is the only triggering signal a skill has.
Signed-off-by: Konstantin Tursunov <Konstantin.Tursunov@constructor.tech>
The rule says to parse a downloaded export and compare it against the grid it came from. It did not say what does the parsing, so every journey that follows it would arrive at the same reader independently. Name `downloads.py`: `download_export` saves and parses the file, `claimed_row_count` reads the `aria-rowcount` a virtualized grid uses to declare the row total it is only showing a window of, and one normalizer spans CSV strings, XLSX numbers and rendered DOM text so `29` and `29.0` compare equal while an empty cell stays distinct from a zero. The three journeys that download a file are named as the worked examples. The helper itself lands in the test PR; this is the guidance that sends a reader to it. Signed-off-by: Konstantin Tursunov <Konstantin.Tursunov@constructor.tech>
ktursunov
force-pushed
the
codex/report-testing-skills
branch
from
August 17, 2026 07:27
2bf674c to
7be7371
Compare
ktursunov
changed the title
ktursunov
marked this pull request as ready for review
August 17, 2026 09:24
ktursunov
enabled auto-merge
August 17, 2026 09:35
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
.claude/skills/holds the instructions an AI agent follows when it testssomething in this repo. Three of them now cover a configurable surface — a screen
whose options change what comes back, the report builder (#2508) being the
current example. Nothing covered that before, so testing one meant trying a few
combinations and calling it done.
Which skill changed
scope-feature-testsdrive-uistand-ui-testThe rules they now carry
pairwise rather than as a full cross.
event saying something was produced is not evidence about its contents.
reload, Back/Forward — because the validators behind them do not all fail the
same way.
for what only a browser can show.
Also here: which origin to open for the test stand
drive-uinow says to open the seeded compose stand on its gateway port, notthe published frontend port. That port serves the SPA and nothing else, so
/auth/*comes back asindex.htmland sign-in loops — which reads as a brokenfrontend and is not one.
Order
#2597 applies these rules to the browser tests that download a file, and adds the
tests/stand/ui/downloads.pythis guidance names. Merge that one first and thisreads against code that exists.