agentsclimarketplace

Android ui states validation

Skill krutikJain/android-agent-skills/skills/android-ui-states-validation

Android skills repository for Kotlin, Compose, XML, testing, CI, release work, and legacy upgrades

Install
npx -y skills add krutikJain/android-agent-skills --skill android-ui-states-validation

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 14 stars14 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Review Android UI flows for empty, loading, error, offline, and edge-case behavior before release.

SKILL.md

4.4 KB, 725 tokens by cl100k_base, as published. Nobody here has run it

Android UI States Validation

When To Use

  • Use this skill when the request is about: validate android ui states, check loading empty error flow android, edge cases in android screen.
  • Primary outcome: Review Android UI flows for empty, loading, error, offline, and edge-case behavior before release.
  • Reach for this skill when the core question is state coverage and recovery UX, not which test framework or assertion library to use.
  • This skill decides what the user should see in each branch of the state matrix. android-testing-ui only validates that design with assertions and screenshots.
  • Handoff skills when the scope expands:
  • android-compose-accessibility
  • android-testing-ui

Workflow

  1. Confirm the user-visible journey, target device behavior, and failure states that matter.
  2. Build a state matrix that covers loading, content, empty, error, offline, stale-data, denied-permission, and post-action confirmation states as applicable.
  3. Identify the owning screens, activities, destinations, and state holders for each branch of that matrix.
  4. Validate recovery UX, accessibility, configuration changes, and back-stack behavior in the showcase apps.
  5. Hand off to testing only when the missing work is about assertions or automation rather than state design.

Guardrails

  • Treat loading, empty, error, offline, and permission-denied states as first-class UI states.
  • Do not hide navigation or permission side effects inside reusable UI components.
  • Prefer lifecycle-aware APIs over manual callback chains.
  • Keep deep links, intents, and permission prompts testable and observable.
  • Distinguish transient messages, persistent error states, and stale-but-usable content instead of collapsing them into one generic failure screen.

Anti-Patterns

  • Assuming the happy path is enough for product flows.
  • Hard-coding request codes or route strings in multiple places.
  • Triggering navigation directly from repositories or network layers.
  • Shipping flows without recovery UI for denied permissions or broken state.
  • Using tests as a substitute for deciding what the product should show in each failure mode.

Examples

Happy path

  • Scenario: Validate OrbitTasks loading, content, and success confirmation states.
  • Command: cd examples/orbittasks-compose && ./gradlew :app:connectedDebugAndroidTest

Edge case

  • Scenario: Exercise long content, empty lists, and sync failures in the XML fixture.
  • Command: cd examples/orbittasks-xml && ./gradlew :app:connectedDebugAndroidTest

Failure recovery

  • Scenario: Avoid misrouting UI validation work to accessibility or testing-only skills.
  • Command: python3 scripts/eval_triggers.py --skill android-ui-states-validation

Done Checklist

  • The implementation path is explicit, minimal, and tied to the right Android surface.
  • Relevant example commands and benchmark prompts have been exercised or updated.
  • Handoffs to adjacent skills are documented when the request crosses boundaries.
  • Official references cover the chosen pattern and the main migration or troubleshooting path.

Official References

What ships with it: 5 files

2.9 KB alongside SKILL.md, 1 of them executable

agents/

scripts/

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.