Debrief
Open-source AI workbench for interview prep: resume + JD to role intel, story bank, scripts, mock interviews, and live copilot.
npx -y skills add YunyueLi/greenroom --skill debriefAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 4 stars4 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Post-interview review — reconstructs the Q&A from transcript or memory, captures interviewer intel and next-round forecasts, and turns the round into a revision list for the script. Use when the user says they just finished an interview, 刚面完 / 复盘一下 / 面试录音转文字给你 / 这轮被问了什么, or pastes an interview transcript or recollection. Do NOT use for practice sessions (mock-interview) or pre-interview research (job-intel).
The file declares its own license as MIT. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.
SKILL.md
2.6 KB, as published. Nobody here has run it
debrief · 面后复盘
每轮真实面试后 24 小时内做一次（记忆衰减很快）。产出 jobs/<slug>/rounds/rN-debrief.md：把这一轮固化成下一轮能用的东西——问答复原、面试官情报、下轮预测、逐字稿修订清单。
输入采集
最好的输入是录音转写。没有就引导回忆倾倒，按这个顺序问（一次 2-3 个问题，别审讯）：
- 从进场开始按时间顺序过：第一个问题是什么？你怎么答的？
- 每个问题：面试官当时的反应（追问了？点头？记笔记？换话题？）
- 哪几题答得不顺、答完自己不满意，或被追问到答不上来？
- 面试官透露了什么：团队情况、岗位实情、下轮安排、他自己的背景
- 反问环节你问了什么、对方答了什么
明确告诉用户：记不清的就说记不清，复原里会标「不确定」，比硬凑可靠。
产出 rN-debrief.md
frontmatter：type: debrief / job: <slug> / round: N / updated: <date>。四个部分：
1. 问答复原
按时间顺序，**Q:** / **A:**（A 记要点和关键原话）。不确定处标 （不确定）。面试官的反应用括号注在对应位置。
2. 面试官情报更新
观察到的风格（追问型？要数字型？）、在意什么（哪个话题追了三层）、他主动透露的信息（团队/方向/他的背景）。回写进 intel.md 的面试官档案。
3. 下轮预测与建议
- 本轮剧透的下轮信息：谁面、考什么（面试官经常顺嘴说"下一轮会让你聊聊 X"——这是最高价值情报）
- 本轮暴露的弱点 → 下轮大概率重点核查的点
- 给 interview-script 的修订清单：哪些题答得不顺要重写、下轮要新增什么题
4. 待办
按优先级列 3-5 条（核实什么、补什么材料、改什么稿、跟猎头确认什么）。
收尾
汇报：复原了几题、面试官情报更新了什么、下轮预测、建议的下一步（通常是按修订清单回 interview-script 改稿）。