Use Case

When reviewing multiple sources for research, you need concise summaries that capture key points, methodology, findings, and relevance. Cowork can process multiple documents and create structured summaries.

Example Prompt

I have 8 research papers in ~/Research/Papers/ for my literature review on sustainable packaging. For each paper, please create a summary including:

- Citation (author, year)
- Research question/objective
- Methodology used
- Key findings (3-5 points)
- Limitations mentioned
- Relevance to my topic (sustainable packaging trends)

Compile all summaries into a single literature_summaries.md file

What Cowork Does

Document Processing: Reads each PDF/document in the folder
Extract Metadata: Identifies authors, publication year, journal
Identify Structure: Locates abstract, methodology, results sections
Summarize Content: Creates concise summaries of each section
Assess Relevance: Evaluates connection to your research topic
Compile Report: Generates organized summary document

Sample Output

# Literature Review Summaries
Topic: Sustainable Packaging Trends

---

## 1. Chen & Park (2023)
**Title**: Biodegradable Polymers in Food Packaging
**Journal**: Packaging Technology & Science

### Research Objective
Evaluate performance of PLA-based packaging vs. traditional plastics

### Methodology
- Comparative laboratory testing
- Consumer acceptance study (n=450)
- 6-month shelf life trials

### Key Findings
1. PLA packaging extended shelf life by 15%
2. Consumer preference 67% for eco-labels
3. Cost premium decreased to 12% from 2022
4. Composting rate improved with new formulations
5. Moisture barrier still 23% below traditional

### Limitations
- Single product category (fresh produce)
- Limited geographic scope

### Relevance: High
Directly addresses biodegradable material trends

---

## 2. Rodriguez et al. (2023)
[Next paper summary...]

Tips

Group related papers in the same folder
Specify your research focus for relevance scoring
Request comparison tables if needed
Ask for gap analysis across sources