android-emulator-automation
Use semantic ADB and UIAutomator workflows to inspect, launch, and interact with Android apps from agents.
47
35%
Does it follow best practices?
Impact
Pending
No eval scenarios have been run
Passed
No known issues
Optimize this skill with Tessl
npx tessl skill review --optimize ./.github/skills/android-emulator-automation/SKILL.mdQuality
Discovery
N/ABased on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
Something went wrong
Implementation
35%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill reads more like a process document or checklist than an actionable automation guide. It describes what to do at a high level but never shows the actual ADB or UIAutomator commands (e.g., `adb shell uiautomator dump`, `adb shell input tap`), making it largely non-actionable. The structure is reasonable but the lack of concrete, executable examples is a significant weakness for a skill that should teach Claude how to perform specific automation tasks.
Suggestions
Add concrete, executable ADB and UIAutomator commands inline (e.g., `adb wait-for-device`, `adb shell uiautomator dump /sdcard/window_dump.xml`, `adb shell input tap X Y`) instead of only referencing external scripts.
Include at least one complete worked example showing the actual command sequence from device verification through UI dump to semantic tap, with expected output.
Add explicit validation commands at each workflow step (e.g., `adb shell getprop sys.boot_completed` should return `1` before proceeding).
Reduce overlap between Guardrails and Anti-Patterns sections — consolidate into a single constraints section to improve conciseness.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The content is reasonably structured but includes some unnecessary sections like 'When To Use' that restate obvious context, and the guardrails/anti-patterns sections overlap conceptually. The workflow steps are described abstractly rather than with concrete commands, adding words without adding value. | 2 / 3 |
Actionability | The skill provides no executable ADB commands, no concrete code snippets for UI hierarchy dumping or tapping, and no actual command-line examples beyond references to external scripts. A reader cannot copy-paste anything to actually perform Android automation — the examples just point to scripts without showing their content or the underlying ADB/UIAutomator commands. | 1 / 3 |
Workflow Clarity | The workflow has a clear 5-step sequence and mentions verification of device state, but lacks explicit validation checkpoints with concrete commands. Steps like 'Dump the current UI hierarchy' and 'Verify the Android SDK' don't show how to do so, and there's no feedback loop for error recovery beyond the abstract 'device_health.py' reference. | 2 / 3 |
Progressive Disclosure | The content references external scripts and handoff skills, showing some progressive disclosure structure. However, the references to scripts (run_examples.sh, navigator.py, device_health.py) are opaque — there's no indication of what they contain or what output to expect, and the handoff skills are listed without clear signaling of when/why to navigate to them. | 2 / 3 |
Total | 7 / 12 Passed |
Validation
90%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 10 / 11 Passed
Validation for skill structure
| Criteria | Description | Result |
|---|---|---|
metadata_field | 'metadata' should map string keys to string values | Warning |
Total | 10 / 11 Passed | |
- Repository
- krutikJain/android-agent-skills
- Commit
c5bf673
- Reviewed
Table of Contents
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.