
Interview Cheat Sheet Writer
Writes long Chinese ML/LLM interview-prep guides with formulas, PyTorch code and 25 practice questions.
- Stars
- 18,482
- Licence
- MIT
- Files
- 1 file
What it does
When you ask for an interview prep guide on a machine learning or LLM topic, the agent follows a strict template to write a 600 to 1000 line Chinese tutorial. It covers intuition, key formulas with derivations, from-scratch PyTorch code, comparison tables, and 25 interview questions graded from basic to advanced, with collapsible answers. It then sends the draft to a second AI model for math and code review before rendering it, and it does not commit files to git unless you approve.
Great for
- Students preparing for ML or LLM engineering interviews who read Chinese
- Researchers who want a quick structured refresher on a specific technique such as RLHF, DPO or MoE
- Authors of Chinese-language technical tutorials who want a consistent house style
Skip it if
- Broad topics like all of LLM training, which the skill itself says are too wide and should be split; output quality drops and it may stop to ask for scoping
- Readers who need English content, since the template is Chinese-first
- Anyone relying on the output unverified: the cross-model review is an automated check and can still miss errors in formulas, citations or code
Try asking
- “写一份 RLHF / DPO / PPO 面试 cheat sheet”
- “帮我准备 KV Cache 的面试题”
- “出一份 MoE 速查,balanced 模式”
- “写一份 speculative decoding 的教程 with --effort max”
Install in 3 steps
- 1
Download the skill
The ZIP contains the 1 original files, the MIT licence and an attribution note.
interview-cheatsheet.zip - 2
Add it to your agent
Claude Code (all projects)
mkdir -p ~/.claude/skills && unzip interview-cheatsheet.zip -d ~/.claude/skills/Claude Code / Agent SDK (this project only)
mkdir -p .claude/skills && unzip interview-cheatsheet.zip -d .claude/skills/Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.
- 3
Ask for the task
No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.
Get better results
- Scope the topic narrowly (for example 'DPO vs IPO' rather than 'alignment') so the output fits the 600 to 1000 line target
- Review the generated formulas and code yourself before publishing, since the automated review is not a substitute for expert checking
- Use --byline at render time for your name and affiliation rather than putting personal details in the body
What's inside
- SKILL.md
Redistributed unmodified from wanshuiyin/Auto-claude-code-research-in-sleep under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.