| name | research-topic |
| description | Research a technical or academic topic in depth to support article writing. Finds key papers, surveys the state of the art, identifies competing approaches, and produces a research brief with citations. Use when preparing to write an article and needing to understand the landscape and find references. |
| argument-hint | <topic> [output_dir] [target_journal] |
| allowed-tools | Read, Write, Glob, Grep, WebSearch, WebFetch, Task |
Topic Research for Article Writing
Research a topic deeply to support writing a technical or academic article, with a focus on what has been published and what gaps remain.
Arguments
$0 — The topic to research (required; can be a short phrase or a question, e.g. "LLM-based code generation for embedded systems" or "does pair programming improve code quality?")
$1 — Output directory (optional; defaults to output/<topic_slug>/)
$2 — Target journal (optional; shapes the depth and style of research — practitioner vs. research)
Workflow
Phase 1: Understand the Topic
Break the topic down into:
- Core concept — What is the subject?
- Key sub-questions — What specific aspects need to be covered?
- Key terms — What search terms will find the best material?
- Competing framings — Are there different schools of thought?
Phase 2: Literature Search
Search for material systematically. For each category below, run 2–4 targeted searches.
2a. Foundational Papers and Surveys
Search: <topic> survey, <topic> review, <topic> overview
Look for papers with >100 citations or from top venues (ICSE, FSE, ISSTA, PLDI, OOPSLA, etc.)
2b. Recent Work (last 3 years)
Search: <topic> 2024, <topic> 2025, <topic> 2026
Focus on novel results, tools, datasets, benchmarks.
2c. Practical/Industry Angle
Search: <topic> industry, <topic> practice, <topic> case study
Especially important for IEEE Software and CACM Practice submissions.
2d. Contrarian or Critical Views
Search: <topic> limitations, <topic> criticism, <topic> failure, <topic> challenges
A strong article acknowledges the limits of the approach.
2e. Tools, Benchmarks, Datasets
Search: <topic> tool, <topic> benchmark, <topic> dataset, <topic> open source
Phase 3: Read and Extract
For the top 5–10 most relevant sources found, use WebFetch to read abstracts and key sections.
Extract for each source:
- Title, authors, year, venue
- Key claim / contribution
- Methodology (if applicable)
- Key findings / numbers
- Limitations
- Relation to the target article topic
Phase 4: Identify the Gap
After surveying the landscape, identify:
- What is well-established — broadly accepted, widely replicated
- What is actively debated — multiple competing approaches, no consensus
- What is missing — topics not covered, approaches not tried, questions not asked
- Where the article fits — how the author's contribution relates to the above
Phase 5: Produce Research Brief
Write a research brief to output/<topic_slug>/<topic_slug>_research_brief.md with this structure:
# Research Brief: <Topic>
**Prepared:** <date>
**Target journal:** <journal or "General">
**Topic query:** <original topic phrase>
---
## 1. Topic Overview
<2-3 paragraphs explaining the topic, why it matters, and how it is typically studied>
## 2. Key Concepts and Terminology
| Term | Definition | First use |
|------|-----------|-----------|
| ... | ... | ... |
## 3. Foundational Work
| Paper/Work | Authors | Year | Contribution | Venue |
|-----------|---------|------|-------------|-------|
| ... | ... | ... | ... | ... |
## 4. State of the Art (Recent Work)
| Paper/Work | Authors | Year | Contribution | Key Result |
|-----------|---------|------|-------------|-----------|
| ... | ... | ... | ... | ... |
## 5. Industry / Practitioner Perspective
<What practitioners know/believe about this topic; what tools exist; what problems remain unsolved in practice>
## 6. Key Debates and Open Questions
(What is contested? What has been tried and failed?)
## 7. The Gap — Where New Work Fits
<What is missing that a new article could contribute>
## 8. Suggested Related Work to Cite
(Organized by section/argument of a potential article)
## 9. Sources
| # | Title | URL / Reference | Year | Notes |
|---|-------|----------------|------|-------|
| ... | ... | ... | ... | ... |
Phase 6: (Optional) Search for Competing Submissions
If time permits, search for recent arXiv preprints or workshop papers on this topic to check whether very similar work has already been published recently:
- Search:
arxiv <topic> 2025, arxiv <topic> 2026
Output
output/<topic_slug>/<topic_slug>_research_brief.md — Research brief with structured findings and citations
Notes
- Prioritize papers from top-tier venues (ICSE, FSE, PLDI, SOSP, etc.) and IEEE/ACM publications
- For IEEE Software and CACM Practice, spend more time on the industry/practitioner angle; these editors expect work grounded in real-world experience
- For IEEE TSE, cover methodology and related work in depth — reviewers expect comprehensive literature awareness
- If the user's core claim is already well-established in the literature, flag this prominently — it means the framing or contribution needs sharpening
- Use the
data/journal_profiles.json file in this project to understand what each journal values