TwotandAlex Rudenko bf8c6d9bc7 docs(skills): clarify Lighthouse score scale and contrast workflow (#3006) (#3218) ...

Fixes #3006.

In `skills/a11y-debugging/SKILL.md`:
- Clarify in Step 1.2 ("Analyze the Summary") that `lighthouse_audit`
default text output formats category scores on a 0-100 scale followed by
`Passed: N` and `Failed: N` audit counts. The 0-1 scale
(`summary.scores[].score`) and `summary.audits.failed` are present in
structured content when `--experimentalStructuredContent` is used.
- In Step 7 ("Color Contrast"), direct agents to use the "Check Color
Contrast" snippet via `evaluate_script` directly, clarifying that
`list_console_messages` with `types: ["issue"]` only surfaces issues
already emitted by Chrome and does not request a contrast check.

## Testing

- Ran `npx prettier --check skills/a11y-debugging/SKILL.md` (passes
cleanly).
- Verified `npm run build` and `npm run typecheck` (`tsc --noEmit`) pass
with 0 errors.

---------

Co-authored-by: Alex Rudenko <alexrudenko@chromium.org>

2026-10-20 17:11:59 +00:00

..

references

fix: repair broken markdown and extract snippets in a11y-debugging skill (#1096)

2026-03-04 08:05:47 +00:00

SKILL.md

docs(skills): clarify Lighthouse score scale and contrast workflow (#3006) (#3218)

2026-10-20 17:11:59 +00:00