sc-ux
Skill13 starsChanged 2 months agoMITView on GitHub
SI-Coder user experience skill for usable, accessible, predictable web interactions: navigation, information architecture, feedback, keyboard behavior, errors/recovery, forms, loading/empty states, responsive task priority, and desktop workbench ergonomics.
In this repository28 other files it gives its agents.
- AGENTS.mdAGENTS.md
- si-coderSkill
- sc-allSkill
- sc-axSkill
- sc-buildSkill
- sc-cfSkill
- sc-clerkSkill
- sc-convex-cloudSkill
- sc-convexSkill
- sc-dokploySkill
- sc-dokuSkill
- sc-dxSkill
- sc-feSkill
- sc-gitSkill
- sc-helpSkill
- sc-installSkill
- sc-n8nSkill
- sc-onboardingSkill
- sc-providerSkill
- sc-resendSkill
- scSkill
- sc-skillSkill
- sc-stripeSkill
- sc-supabaseSkill
- sc-syncSkill
- sc-uiSkill
- sc-vercelSkill
- use-si-coderSkill
The file
We split this skill into the rules it gives an agent
9 rules. About 0 are common, 9 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈9
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 15
Keep durable instructions in English.
- Distinctive: line 20
Before changing a flow, identify:
- Distinctive: line 33
Every important flow should satisfy:
- Distinctive: line 39
Prefer recognition over memory: labels, context, recent choices, and visible affordances beat hidden recall.
- Distinctive: line 52
Do not use ambiguous spinners for long operations when meaningful progress/state can be shown.
- Distinctive: line 59
keep global navigation, local navigation, work area, and contextual tools conceptually distinct,
- Distinctive: line 61
use one tab stop for composite toolbars/lists where appropriate and arrow-key navigation inside them,
- Distinctive: line 65
keep primary work visible.
- Distinctive: line 72
When tools allow, test the real flow rather than only reading components:
≈ “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
sc-ux — usable before impressive
Language
Keep durable instructions in English. Reply in the user's language unless requested otherwise.
Core model
Optimize for task completion, comprehension, control, recovery, and accessibility. Do not equate fewer visible controls with better UX; hide complexity only when users can still discover, understand, and recover from actions.
Before changing a flow, identify:
- primary user jobs and frequency,
- user expertise and likely mental model,
- entry points and navigation path,
- current state model,
- irreversible/destructive actions,
- latency and failure points,
- keyboard/touch/pointer requirements,
- accessibility requirements,
- responsive task priority,
- explicit scope exclusions.
Usability gates
Every important flow should satisfy:
- Status is visible: users can tell what is happening, what changed, and what is selected.
- Language matches the user's task, not internal architecture.
- Users retain control: cancel/back/undo/retry where meaningful.
- Patterns are consistent across the product unless a difference communicates meaning.
- Prevent errors before explaining them.
- Prefer recognition over memory: labels, context, recent choices, and visible affordances beat hidden recall.
- Frequent tasks are efficient without making first use opaque.
- Interfaces contain only information/actions that earn their place.
- Errors explain impact and next recovery action.
- Help appears near the blocking context rather than as a documentation dump.
Interaction state contract
Important async/action surfaces must cover applicable states:
idle → pending → progress → success | error → retry/cancel.
Data surfaces must cover applicable states:
loading → populated | empty | error | permission-denied | offline/stale.
Do not use ambiguous spinners for long operations when meaningful progress/state can be shown.
Accessibility baseline
Use semantic controls and names. Preserve visible focus. Keyboard users must reach and operate every primary action without pointer-only traps. Do not encode meaning by color alone. Maintain sufficient text/control contrast. Respect reduced motion. Provide non-drag alternatives for required drag interactions. Make touch/pointer targets practical and avoid destructive actions adjacent to frequent controls without separation/confirmation/undo.
Desktop/workbench UX
For desktop shells and productivity tools:
- keep global navigation, local navigation, work area, and contextual tools conceptually distinct,
- support predictable focus movement between regions,
- use one tab stop for composite toolbars/lists where appropriate and arrow-key navigation inside them,
- preserve user context when opening/closing inspectors, dialogs, and secondary panes,
- make resize/collapse boundaries discoverable,
- remember reasonable layout state when product requirements allow,
- keep primary work visible; do not let chrome dominate the canvas,
- command palettes/shortcuts accelerate frequent actions but do not replace discoverable UI for essential tasks.
Responsive behavior
Responsive UX is reprioritization, not desktop columns → one long mobile stack. Preserve the highest-value task and state. Do not alter an explicitly excluded mobile/navigation surface.
Verification
When tools allow, test the real flow rather than only reading components:
- first-use path,
- frequent-user path,
- keyboard-only path,
- error/retry path,
- empty/loading path,
- destructive action recovery,
- representative viewport resize,
- deep link/back-forward behavior where relevant.
Record defects as problem → user impact → fix → verification.
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.

