OpenGithubs / OpenGithubs/weekly

【开源自荐】character-card-skills:让 Agent 写出「活人感」AI 角色卡(附盲评实验:LLM 评委判对率只有 12%)

Open
#134 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
No language data
Stars
1.7k
Forks
71
PR merge metrics
No merged PRs in 30d

Description

1). 项目名称:character-card-skills

2). 项目地址:https://github.com/foreverse-app/character-card-skills

3). 项目简介:
给 Claude Code / Cursor / Codex / Gemini CLI 用的 AI 角色卡创作技能包(SKILL.md 约定):人设四轴选型、开场白五范式、世界书套路、卡面文案句式纪律,附 18 张可直接导入 SillyTavern 的原创示范卡(v2 JSON + v3 PNG)和一个校准过的中文 AI 味检测器。代码 MIT,内容 CC BY 4.0。

4). 推荐原因:

  • 一个反直觉的实验:我们让 LLM 评委盲评「人写的卡 vs AI 写的卡」,评委间一致率 86%,判对率只有 12%——模型把细节密度、状态栏排版当「人味」,真人读者把同样的东西当机器指纹。完整记录在仓库 docs/blind-review-experiment.md
  • 检测器按句式病灶打分(引语点题、对仗翻转、伪精确数字),拿人类卡和 AI 卡做过 gold 校准,人类样本零误杀,可以当 CI 闸门用(仓库自己的 cards gate 就在跑)
  • 英文原生 skill:据我们所知是 agent-skills 生态第一个英文原生的写卡 skill,校准来自英文社区 30 张高星卡的实卡采样,不是中文规则机翻
  • 18 张示范卡全部过「检测器 + 人工」双闸,导入 SillyTavern / Foreverse 直接能玩

5). 项目图片:
(README 有示例)

分类:AI / Agent Skills / 角色扮演 / SillyTavern

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by reviewing the linked character-card-skills repository and its docs/blind-review-experiment.md. This issue names no target file, test, or requested change in the weekly repository, so a concrete definition of done cannot be determined from the text alone.

Written by the indexing model from the issue text.

Assessment

Domain
content
Issue type
Documentation
Difficulty
5/5
Estimated time
Over a week
Activity status
Quiet
Clarity
Needs clarification
Newbie friendliness
20/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.