Cyber Emperor - review filter for steam

This is an LLM-powered Steam review management tool designed to help indie game developers obtain valuable player feedback without being directly exposed to hostile reviews.

For creators, the motivation and joy brought by a hundred positive reviews can be completely offset by a single hostile comment.

This tool uses an LLM of your choice (you will need an API key) to categorize each review (positive, constructive suggestion, contains a bug report, unfriendly tone, purely malicious), and also adds a brief summary of the content. This allows you to know what you are about to read before opening a review, and decide whether you really want to click on it.

You cannot stop hostile reviews from being posted, but you can protect yourself from being hurt — don’t read them, don’t reply to them.

Additionally, inspired by SillyTavern, the tool includes extensive role-playing elements. You can define a world setting and your own identity, and assign a specific character to classify reviews for you.

— Feel free to share your custom character cards! (Please refer to the separately distributed narrator_sample for the format.)

LLM 驱动的 Steam 评论管理工具，帮助独立游戏开发者在不直接暴露于恶意评论的情况下获取有价值的玩家反馈。

对创作者来说，一百篇好评带来的动力和快乐可能会被一篇充满恶意的评论抵消。

这个工具可以使用你自己选择的LLM（你需要一个API Key）为每个评论分类（好评、善意建议、包含BUG报告、态度不友好、纯粹恶意）、并追加了内容简述，让你在阅读评论之前先知道自己要面对什么，以及考虑好是不是真的要点进去。

你无法阻止恶意评论的发布，但是可以保护自己不被伤害——不要看，不要回答。

另外，受到SillyTavern的启发，加入了大量角色扮演要素，你可以定义世界观和自己的身份，并让指定的角色为你进行评论分类。

——欢迎分享你自制的角色卡！（格式请参考单独分发的narrator_sample）