ui-experience
Skill0 starsChanged 22 days agoMITView on GitHub
Design or review screens, interaction flows, component states, visual hierarchy, motion, responsiveness, localisation and accessibility. Use it for UI proposals, prototypes and UI changes; don't use it for pure business-logic changes.
In this repository7 other files it gives its agents.
- architecture-changeSkill
- change-reviewSkill
- idea-to-planSkill
- orchestrateSkill
- product-changeSkill
- quality-testingSkill
- start-projectSkill
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 4 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈4
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 12
When the task touches product boundaries, remote capabilities or high-impact actions, read the necessary parts of the project rules and the …
- Distinctive: line 13
When you need the target screen, look up the matching screen in the prototype document (screen numbers like P01) and separate what exists, …
- Distinctive: line 14
Look at the existing theme, shared components and neighbouring screens and keep the established design language.
- Distinctive: line 33
Check small screens, enlarged system fonts, the keyboard, safe areas, scrolling, different aspect ratios and touch-target sizes.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
UI and interaction
Make the interface represent what the product can really do, and keep it clear, operable and consistent across states, screen sizes and input conditions. If the global rules say the user has no design training, propose 2–3 visual directions with your reasoning and ask the user to choose.
Get the context
- When the task touches product boundaries, remote capabilities or high-impact actions, read the necessary parts of the project rules and the PRD.
- When you need the target screen, look up the matching screen in the prototype document (screen numbers like P01) and separate what exists, what is targeted and what is a placeholder.
- Look at the existing theme, shared components and neighbouring screens and keep the established design language; if the project has a design-system document, it wins.
- Confirm the current UI from code and tests; don't infer from the prototype that something is implemented.
Flows and states
First settle the user's goal, entry point, primary action, information priority, way back and completion feedback. According to the risk of the feature, check:
- Normal, loading, empty, failed and retry.
- Offline, disabled, permission denied and external capability unavailable.
- Progress, cancellation, interruption and recovery for long tasks.
- Repeated taps, quick switching, going back, the app moving to the background and data refreshes.
- Explanation and confirmation for high-impact actions such as delete, overwrite, purchase and upload.
Remote, payment or experimental capabilities that aren't connected must never appear successful; either hide the entry or show a real disabled/unavailable state that explains the condition.
Visuals and accessibility
- Express priority with hierarchy, spacing, alignment and consistent action placement rather than piling on decoration.
- Compose existing components first; add a new component only when it will really be reused.
- Check small screens, enlarged system fonts, the keyboard, safe areas, scrolling, different aspect ratios and touch-target sizes.
- Colour, sound and motion must never be the only carrier of information; key controls need semantic labels and understandable reasons when disabled.
- Motion respects the system's reduce-motion setting and avoids rapid flashing; animation callbacks must not be the only source of business state.
- User-visible text is managed centrally (using the project's localisation approach), with attention to text length and wrapping across languages.
Implementation boundaries
The UI layer only renders state and forwards actions. It doesn't access databases, files, the network or payments directly, parse protocols, or make business decisions.
Output
Scaled to the task, describe: the user path, relevant states, reused or new components, handling of responsiveness/localisation/accessibility, how it is verified (component tests and manual visual checks), and any differences from the PRD or prototype. When the user only asked for a UI proposal or prototype, don't change business logic.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

