Colour Contrast Guardrails
Stop custom colours and themes from making text and controls unreadable.
What it adds
An audit of the app's colour tokens against WCAG contrast thresholds, fixes made in the palette itself, and an automated check that keeps regressions out.
What your agent is told to do
6
What your agent is told to do
6-
1
Start by auditing the existing tokens: compute the contrast ratio for every foreground and background pair the app actually uses, and report the failures before changing anything.
-
2
Fix the failures in the palette, not in the components that happen to expose them. Patching one button leaves the same bad pair waiting in nine other places.
-
3
Check the combinations people forget: disabled text, placeholder text, borders on inputs, icons carrying meaning, and text sitting over images or gradients.
-
4
Check focus indicators against both the component and the surface behind it. A focus ring that only contrasts with one of them disappears half the time.
-
5
Add a check that runs in CI over the token pairs and fails the build on a regression, so a designer adding a lighter grey next quarter finds out immediately.
-
6
Do NOT silently darken a brand colour to make it pass. Report the conflict and propose an adjusted value or a different pairing; changing brand colours without asking is not yours to do.
Edge cases it handles
7
Edge cases it handles
7- Opacity changes the effective colour — compute against the composited result, not the declared value.
- Large text has a lower threshold, so an audit that applies one number everywhere will report failures that are not failures.
- Text over photographs has no single background; require a scrim or a solid plate rather than guessing an average.
- Every interactive state needs checking — hover, active, selected, visited, disabled — not just the resting one.
- Dark mode needs its own pass. Inverted palettes routinely fail where the light theme passed.
- Charts and status colours must not rely on hue alone; two hues at equal lightness are indistinguishable to many users regardless of contrast.
- If users can pick their own accent colour, validate it at the point of selection instead of shipping a broken theme.
Definition of done
8
Definition of done
8- Every foreground and background pair in use has been measured and the failures reported.
- Body text meets WCAG AA, and non-text UI (borders, icons, focus rings) meets 3:1.
- Fixes were made in the colour tokens rather than in individual components.
- Disabled text, placeholder text, and text over images all pass or are explicitly documented as exempt.
- Contrast is verified in every theme and every interactive state.
- A CI check fails the build when a new token pair drops below threshold.
- The feature matches the existing design system.
- No existing functionality is broken.
Related features
Auto Captions
Auto Captions
Every uploaded video gets timed captions without anyone typing them out.
What it does
An automatic captioning pass over uploaded video that produces a timed caption track shown in the player.
How it works
- 1 Extend the app's existing Voice Transcription rather than adding a second speech pipeline. What is new here is timing, storage as a caption track, and player integration, not the recognition itself.
- 2 Run captioning as a background job triggered after the upload finishes, and reuse the app's existing job progress reporting so the user can see where a long file has got to.
- 3 Store the caption track as its own record linked to the media, with a language, a status, and a source marking it as machine-generated so a human correction can supersede it later.
Copy the prompt
No account needed
Add this feature to my app:
https://addthisfeature.com/x/auto-captions
Accessible Name Validator
Accessible Name Validator
Find every control a screen reader announces as nothing useful.
What it does
An audit of the computed accessible name of each interactive element in the rendered interface.
How it works
- 1 Compute the name the way assistive technology does — following the resolution order through labels, referenced elements, attributes, and text content — rather than reading the source markup and guessing.
- 2 Report three separate failures: an empty name, a generic name such as button, link, or here, and duplicate names within the same context where several controls announce identically.
- 3 Prefer a fix that adds a visible label. A control given a hidden name is announced but still mysterious to the sighted user who cannot tell two identical icons apart.
Copy the prompt
No account needed
Add this feature to my app:
https://addthisfeature.com/x/accessible-name-validator
Touch Target Checker
Touch Target Checker
Find the controls that are too small or too crowded to hit reliably on a phone.
What it does
A development-time audit reporting the measured hit area and neighbour spacing of every interactive element.
How it works
- 1 Measure the element's actual activation area — the region that responds to a tap — rather than the icon or glyph drawn inside it, since padding and pseudo-element extensions frequently make a small icon a perfectly adequate target.
- 2 Check the gap between adjacent targets as well as their size. Two comfortable buttons sitting flush against each other still produce mis-taps, and the report must say which of the two problems it found.
- 3 Report the offending elements with enough context to locate them — the route, the component, and a way to highlight the element on the page — rather than a count of failures.
Copy the prompt
No account needed
Add this feature to my app:
https://addthisfeature.com/x/touch-target-checker
How it works
-
1
Copy the link
Grab the Markdown instruction URL for this feature.
-
2
Give it to your AI
Paste it into Claude Code, Cursor, v0, Lovable — whatever you build with.
-
3
It inspects, then implements
Your agent reads your existing app first, then adds the feature to fit it.
Works with your stack
These instructions are written to adapt. They tell the agent to detect your framework, match your existing design system, and reuse what you already have — rather than assuming a particular stack.
Need it tighter than that? Customize the feature and tell it exactly what you're running.