Lab Report Writer
A Claude Code Skill for generating professional lab reports โ university coursework, scientific research, engineering tests, and medical/material experiments.
Features
- Three input modes: From topic โ full report ยท From raw data โ analysis + report ยท From draft โ improvement & formatting
- Four report types: University lab ยท Scientific research ยท Engineering test ยท Medical/material
- Auto-research: WebSearch for theoretical background and real citable references
- Data processing: Statistics, uncertainty propagation, least-squares fitting
- Charts: Line / scatter / bar with error bars (matplotlib)
- Low plagiarism: Original writing, structure variation, no copy-paste
- Multiple output formats: docx ยท markdown ยท both
- Grading-aligned: Reads rubric files and aligns content to maximize score
Installation
git clone https://github.com/dxkjuanjuan/lab-report-writer.git ~/.claude/skills/lab-report-writer
Then restart Claude Code.
Usage
Just describe what you need in natural language:
ๅธฎๆๅไธไปฝๅคงๅญฆ็ฉ็ๅฎ้ชๆฅๅ๏ผ้ข็ฎๆฏ"็จๅๆๆต้ๅๅ ้ๅบฆ"๏ผ5000ๅญๅทฆๅณ๏ผ่พๅบdocx
or
Generate a lab report from this CSV data: ./data.csv โ fit a linear regression, output as docx
The skill will ask clarifying questions, then execute:
research โ write โ data analysis โ generate doc โ self-check โ deliver
Structure
lab-report-writer/
โโโ SKILL.md # Main skill instructions
โโโ README.md # This file
โโโ LICENSE # MIT
โโโ scripts/
โ โโโ generate-lab-report.js # docx generation (Node.js + docx)
โ โโโ lab_data_analysis.py # Data analysis & charts (Python + matplotlib)
โโโ resources/
โโโ templates/
โโโ university-lab-report-template.md
Prerequisites
| Tool | Purpose | Required |
|---|---|---|
| Claude Code 2.0+ | Run the skill | Yes |
Node.js 18+ + docx npm |
Generate Word documents | For docx output |
| Python 3 + numpy + matplotlib | Data analysis & charts | For data mode |
Install docx globally:
npm install -g docx
Report Types
| Type | Sections |
|---|---|
| University lab | Abstract ยท Intro ยท Principles ยท Apparatus ยท Procedure ยท Data ยท Analysis ยท Conclusion ยท Reflection ยท References |
| Scientific research | Abstract ยท Intro ยท Methods ยท Results ยท Discussion ยท Conclusion ยท References |
| Engineering test | Objective ยท Method ยท Environment ยท Data ยท Analysis ยท Conclusions ยท Appendices |
| Medical/material | Abstract ยท Intro ยท Materials ยท Methods ยท Results ยท Discussion ยท Conclusion |
Low-Plagiarism Strategy
- Original phrasing โ never copy source sentences
- Structure variation โ don't follow any single source's paragraph order
- Specific details โ concrete numbers and technical terms reduce similarity
- Personal analysis โ independent reasoning in the discussion section
- Chinese-English mixing โ technical terms naturally lower similarity scores
Reference Format
Follows GB/T 7714 by default. Supports [J] journals, [M] books, [C] conferences, [R] reports, [Z] standards, [EB/OL] online.
Customization
- New template: Create
resources/templates/<name>.md, reference in SKILL.md - Output format: Edit
scripts/generate-lab-report.jsโ fonts, sizes, margins - Data analysis: Edit
scripts/lab_data_analysis.pyโ new stats or chart types
License
MIT โ see LICENSE.
Created 2026-07-25 ยท Maintainer @dxkjuanjuan ยท Version 1.0.0
No comments yet
Be the first to share your take.