Easybangumi UI Test
Testing & Qualityeasybangumi-ui-testEasybangumi UI Test is an Agent Skill for testing & quality published in the easybangumiorg/EasyBangumi repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.
Use when testing Pure Pure Bangumi (EasyBangumi) Android features through real adb-driven UI flows, including app launch, app data reset, local mock HTTP servers on Mac, device-to-host port mapping, UI hierarchy dumps, reusable page navigation, and evidence collection for manual or scripted feature verification.
How to use this skill
Bring this guide into your coding agent with a prompt tailored to the tool you use.
- Open your project in Codex.
- Copy and paste the prompt below.
- Review files and risks before approving any action.
Inspect the Agent Skill at https://github.com/easybangumiorg/EasyBangumi/blob/HEAD/skills/easybangumi-ui-test/SKILL.md and all supporting files. Report provenance, permissions, scripts, dependencies, and security risks. Propose an exact installation into .agents/skills/easybangumi-ui-test, but wait for my approval before writing files or running scripts.
Copying this prompt does not install or run the skill. Review third-party files before use.
Easybangumi Ui Test
Overview
This skill standardizes real-device testing for Pure Pure Bangumi with adb, local mock services, and repeatable UI evidence collection. Use it when the task is broader than one page or when you need reusable setup, navigation, and validation steps.
Quick Start
- Run the common helper script from the repo root:
./skills/easybangumi-ui-test/scripts/easybangumi_test.sh - If you need a local mock server, start it with:
./skills/easybangumi-ui-test/scripts/easybangumi_test.sh serve /path/to/mock-dir 18080 - If the Android device must access a Mac-local service through
127.0.0.1, use:./skills/easybangumi-ui-test/scripts/easybangumi_test.sh reverse 18080 - Launch and inspect the app:
./skills/easybangumi-ui-test/scripts/easybangumi_test.sh clear com.heyanle.easybangumi4.debug./skills/easybangumi-ui-test/scripts/easybangumi_test.sh launch com.heyanle.easybangumi4.debug./skills/easybangumi-ui-test/scripts/easybangumi_test.sh dump-ui
Workflow
1. Preflight
- Confirm exactly which package to test: usually
com.heyanle.easybangumi4.debug. - Prefer the current workspace build when validating current code.
- Decide whether the test needs clean state:
clearif onboarding, config, or repository state matters. - Decide whether the test needs a mock service:
use
reverse, notforward, when the app on the device should reach a Mac-local service via127.0.0.1.
2. Execute the Real Flow
- Prefer explicit
am/monkeylaunch over tapping a launcher icon when multiple app variants exist. - Use
dump-uibefore important actions and after each state change that matters. - Drive the UI with
tap,text, andswipe. - If onboarding or permissions interrupt the target flow, either complete them manually or document the exact blocking screen before continuing.
3. Validate Outcomes
- Validate both device-side behavior and host-side evidence.
- For network-backed flows, check the local mock server logs for requested paths.
- For install/import flows, verify both the final dialog/toast state and the downstream page state after dismissing dialogs.
- Record unexpected behavior as: trigger, observed result, expected result, likely code area.
4. Clean Up
- Stop any local mock server started for the test.
- Remove temporary
adb reversemappings withunreverse. - Keep test artifacts only if they are useful for follow-up debugging.
Navigation
Read references/navigation.md when the task depends on page entry paths, common bottom-tab routes, or source-management navigation.
Test Strategy
Read references/test-strategy.md when you need a reusable checklist for scenario design, boundary analysis, or result reporting.
Script
Use scripts/easybangumi_test.sh for common setup and adb actions. Prefer the script over retyping raw commands when the action is generic and repeatable.
Supported subcommands:
serve <dir> [port]stop-serve [port]reverse <port>unreverse <port>clear <package>launch <package>dump-ui [output-file]tap <x> <y>text <value>swipe <x1> <y1> <x2> <y2> [duration-ms]screenshot [output-file]current-focus
When To Escalate
- If multiple installed packages share the same app name and launcher icon.
- If the flow requires manual login, device-level permissions, or document picker interaction that cannot be inferred safely.
- If observed UI text conflicts with expected code behavior and you need to decide whether to trust code or runtime.