ワンクリックでManusで任意のスキルを実行

feather-qa-playtester

スター35

フォーク1

更新日2026年6月20日 17:15

Use when an agent should act as a QA engineer for a Feather-enabled Love2D game using sessions, replay, logs, debugger state, plugins, and platform builds.

インストール

Codex または Claude でインストールこの Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。

Manusで実行

ソース

Kyonru

Kyonru/feather

GitHub リポジトリを開く Creator のリポジトリを見る

ダウンロード

Manusで実行

Feather QA Playtester

Workflow

Define the test slice:
- State platform, build command, scene, expected behavior, and pass/fail signals.
Collect evidence:
- Use MCP sessions, logs, observers, performance, debugger state, plugin payloads, and session replay.
- Use platform/device automation only when such tools are actually available in the environment.
Reproduce:
- Prefer Session Replay for repeatable behavior.
- Add breakpoints or observers only around the suspected failure path.
Report like QA:
- Include steps, expected result, actual result, severity, session/replay id, first bad log/state, and proposed next verification.

Bias

Look for user-visible regressions first: crashes, hangs, broken input, incorrect visuals, audio issues, save/load loss, performance spikes, and release-blocking configuration.

References

Read references/workflow.md for test charters, evidence gathering, severity, bug reports, and regression verification.

このリポジトリの他の Skills

同じリポジトリ

feather-debug-builds

Kyonru/feather

Use when an agent should prepare or run debug-oriented Feather builds for desktop, web, Android, iOS, or Steam Deck development.

2026-06-2035

feather-logs-observability

Kyonru/feather

Use when an agent needs to diagnose Love2D behavior from Feather logs, observer values, assets, runtime messages, or telemetry.

2026-06-2035

feather-mcp-live-sessions

Kyonru/feather

Use when an agent should inspect or control a running Feather desktop session through MCP.

2026-06-2035

feather-particle-effects

Kyonru/feather

Use when an agent should create, preview, tune, export, or QA particle systems with Feather MCP Particle Playground tools.

2026-06-2035

feather-performance-profiling

Kyonru/feather

Use when an agent needs to profile frame time, memory, runtime overhead, hotspots, or performance regressions in a Feather session.

2026-06-2035

feather-plugin-iteration

Kyonru/feather

Use when an agent should inspect, configure, enable, or invoke Feather built-in or live-session plugin behavior through MCP.

2026-06-2035

name	feather-qa-playtester
description	Use when an agent should act as a QA engineer for a Feather-enabled Love2D game using sessions, replay, logs, debugger state, plugins, and platform builds.

Feather QA Playtester

Workflow

Define the test slice:
- State platform, build command, scene, expected behavior, and pass/fail signals.
Collect evidence:
- Use MCP sessions, logs, observers, performance, debugger state, plugin payloads, and session replay.
- Use platform/device automation only when such tools are actually available in the environment.
Reproduce:
- Prefer Session Replay for repeatable behavior.
- Add breakpoints or observers only around the suspected failure path.
Report like QA:
- Include steps, expected result, actual result, severity, session/replay id, first bad log/state, and proposed next verification.

Bias

Look for user-visible regressions first: crashes, hangs, broken input, incorrect visuals, audio issues, save/load loss, performance spikes, and release-blocking configuration.

References

Read references/workflow.md for test charters, evidence gathering, severity, bug reports, and regression verification.