draft
Writes a complete IMRaD-structured academic paper draft (Abstract, Introduction, Related Work, Methods, Results, Discussion, Conclusion, References) from the user's notes, documents, and wiki content. Use when the user asks to write a paper, create a manuscript, generate a draft,
What it does
Paper Draft Generation
Generate a complete IMRaD-structured academic paper draft from research notes and documents.
IMRaD Structure
| Section | Purpose | Length |
|---|---|---|
| Abstract | Background, objective, methods, findings, conclusions | 250–300 words |
| Introduction | Problem, motivation, gap, contribution | 400–600 words |
| Related Work | Prior literature organized by theme | 500–800 words |
| Methods | How the research was/will be conducted | 400–600 words |
| Results | What was found (or projected findings) | 300–500 words |
| Discussion | Interpretation, implications, limitations | 400–600 words |
| Conclusion | Summary, contributions, future work | 200–300 words |
| References | All cited sources | As needed |
Workflow
- Gather content — read all provided notes, wiki pages, and documents
- Identify the core contribution — what is new and why does it matter?
- Map citations — track which sources support which claims
- Draft section by section — follow IMRaD order
- Add inline citations — use
[Author, Year]or[Source Title]format - Flag gaps — mark sections needing real data with
> ⚠️ [PROJECTED — add empirical data]
Citation format
Use the actual title or author of notes/documents you read:
Predictive coding theory proposes... [Rao & Ballard, 1999]
At the end, add a References section listing all cited sources.
Handling missing data
When the research doesn't have empirical results yet, write projected results based on the hypotheses — but clearly mark them:
> ⚠️ **Projected Results** — Replace with actual experimental data before submission.
Based on the experimental design, we expect...
Quality checklist
- Abstract covers all 5 components (background, objective, methods, findings, conclusions)
- Introduction ends with a clear statement of contribution
- Every major claim has an inline citation
- Methods section is specific enough to reproduce
- Discussion addresses alternative interpretations
- Limitations section is honest, not defensive
Capabilities
Install
Quality
deterministic score 0.47 from registry signals: · indexed on github topic:agent-skills · 37 github stars · SKILL.md body (2,163 chars)