coder-style-test
Skill2 starsChanged 3 months agoMITView on GitHub
程序员编程风格测试。在对话里出 20 道情景选择题,从起手方式(蓝图/试验)、完成标准(打磨/交付)、和 AI 的分工(亲手/委托)、协作习惯(独行/结伴)四个方面,测出用户在 AI 时代写代码的风格,共 16 种类型,给出强项、要留意的地方、和 Agent 合作的建议,以及可分享的网页版结果链接。触发词包括:"测测我的编程风格"、"程序员风格测试"、"编程人格测试"、"我是什么类型的程序员"、"coder style"、"编程风格测试"。用户只是想评价某段代码的写法或做代码 Review 时,不属于本 skill。
In this repository29 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- ai-hot-pickerSkill
- aihotSkill
- change-meeting-briefSkill
- clarify-before-actionSkill
- dating-chat-coachSkill
- deep-researchSkill
- diagnose-and-explainSkill
- explain-to-masterSkill
- fable-writerSkill
- first-principles-adversarial-reviewSkill
- human-writingSkill
- kakarot-repurposerSkill
- kakarot-visual-explainerSkill
- kakarot-writerSkill
- leaderSkill
- life-designerSkill
- peer-pr-reviewSkill
- ppt-order-workflowSkill
- preschool-lesson-kitSkill
- project-aware-codingSkill
- resume-optimizerSkill
- review-handoffSkill
- steelman-before-answerSkill
- storage-analyzerSkill
- story-video-workflowSkill
- system-structure-diagramSkill
- wechat-publisherSkill
The file
Files
编程风格测试
题目、维度和 16 种类型的文案都在 references/coder-style.json,计分用 scripts/score.py。不要凭记忆出题或改写题目,也不要自己心算结果,一律以这两个文件为准,保证和网页版 https://kakarot-index.kakarot-space.workers.dev/coder-style/ 的结果一致。
流程
1. 出题
读取 references/coder-style.json 的 questions,按原顺序把 20 道题一次性发给用户。每道题写成「题号. 题目」,下面两行是「A. 第一个选项」「B. 第二个选项」,选项顺序和 JSON 里 options 的顺序一致,不能调换。不显示 axis 和 pole 字段,避免用户看出每道题测的是哪一端。
题目前面用一两句话说明:每道题选更接近自己的那一项,没有对错;答完按题号顺序回复 20 个字母,例如 ABBAA BABAB …,中间有空格也行。用户想一道一道地答时,就每次只发一道题,收齐 20 个答案后再计分。
2. 计分
收齐答案后运行:
python3 scripts/score.py "用户的 20 个答案"脚本会检查答案是否正好 20 个 A 或 B,不合格时报错退出。报错时把错误原因告诉用户,请用户补齐或改正,不要自己猜测缺失的答案。用户的回答里夹杂了文字时,按题号整理出 20 个字母再传给脚本,整理结果不确定就向用户确认。
3. 给出结果
把脚本输出的 Markdown 原样发给用户,不改写文案,不额外加对用户性格的评价。最后一行是网页版结果链接,用户打开后能看到带图表的结果页,也可以发给朋友。
用户追问某个类型、某个方面的含义,或者想看全部 16 种类型时,从 JSON 的 axes 和 types 里取内容回答。
边界
- 这是趣味测试,结果仅供参考和娱乐,不用于招聘、绩效等评价场景。用户要求拿它评价别人时,说明这一点。
- 测试只看用户自己的选择,不读取用户的代码、Git 记录或其他文件。
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

