| name | research-paper |
| description | Enterprise-grade autonomous research paper generation skill for AI coding agents — full papers, literature reviews, theses, whitepapers, surveys, policy briefs — with rigorous methodology, statistical validation, multi-style citations (Harvard / APA / IEEE / MLA / Chicago / Nature / arXiv-numeric), and rich visualizations. Activates on slash commands (`/research`, `/paper`, `/literature-review`, `/whitepaper`, `/thesis`, `/survey`, `/policy`) and on natural-language academic-writing requests. Runtime-neutral — works with Claude Code, OpenCode, Cursor, Cline, Codex, Aider, Amp, Antigravity, and 50+ agents via the `npx skills` installer. |
| license | MIT |
| version | 2.4.0 |
Research Paper
A production-grade agent skill that turns any compatible coding agent
(Claude Code, OpenCode, Cursor, Cline, Codex, Aider, Amp, Antigravity,
and 50+ others) into a multi-agent research system:
Orchestrator → Researcher → Methodologist → Analyst → Visualizer
→ Writer → Citation engine → Validator → Reviewer → Publisher.
It produces full, citation-heavy, visually rich, publication-ready outputs in
arXiv / IEEE / ACM / Nature / Harvard styles, plus literature reviews,
theses, technical whitepapers, survey papers, and policy briefs.
This file is the entry point. It is intentionally compact. Heavier
guidance (instructions, workflows, engines, validators, rubrics) lives in
the topic folders below and is loaded on demand via Claude Code's
filesystem tools (progressive disclosure).
1. When to activate
Slash commands (preferred)
| Command | What it does |
|---|
/research <topic> | Full empirical research paper |
/paper <topic> | Same as /research, more permissive |
/literature-review <topic> | Systematic / scoping / narrative literature review |
/whitepaper <topic> | Industry / technical whitepaper |
/thesis <topic> | Thesis / dissertation chapter |
/survey <topic> | State-of-the-art / survey paper |
/policy <topic> | Policy brief or full policy paper |
Common options (any command):
--style [harvard|apa|ieee|mla|chicago|nature|arxiv-numeric],
--format [arxiv|ieee|acm|nature|harvard|...],
--depth [quick|standard|comprehensive],
--sources [N],
--visualizations [auto|N|none],
--audience [academic|technical|executive|general].
Natural-language triggers
- "Write a research paper / academic paper / scientific paper on …"
- "Do a literature review / systematic review on …"
- "Format this draft as IEEE / ACM / arXiv / Nature / Harvard / APA …"
- "Write a thesis chapter / dissertation chapter on …"
- "Produce a whitepaper / survey paper / policy brief on …"
- "Analyze this dataset and write up the findings as a paper."
- "Add citations / bibliography / references in
<style>."
- "Peer-review this draft / validate the methodology."
Do NOT activate for
Blog posts, marketing copy, tweets, casual answers, or single-paragraph
explanations. Those are handled normally without this skill.
2. Operating principles (read every time)
- Anchor to TODAY's date FIRST. Before any planning, search,
or writing, determine today's actual date — via system clock
(
date -u +%Y-%m-%d), runtime context, or asking the user.
Never default silently to the training-data cutoff. Year
ranges (--years last-3) are computed from today, not from
the model's training year. Full protocol:
instructions/freshness.md.
- Plan before writing. Always start with the research plan in
orchestration/pipeline.md. Never jump into prose.
- Progressive disclosure. Only read the file you need for the current
step. Never preload the whole skill.
- Evidence first. Every non-trivial claim is backed by a citation,
dataset, equation, or explicit derivation.
- No hallucinated citations. Never invent DOIs, page numbers,
authors, or volumes. Mark gaps with
[CITATION NEEDED] or
[UNVERIFIED] and surface them in Known gaps.
- Reproducibility. Datasets, code, environments, seeds, and
hyperparameters are documented end-to-end.
- Dual register. Maintain academic rigor and a "Plain-English
summary" for non-specialists (see
prompts/simplification-prompts.md).
- Visual-by-default. Comparisons, trends, distributions, structure,
geography, and processes always get a figure or table
(see
visualization_engine/decision-engine.md).
- Self-review. Run the simulated peer-review pass
(
review_pipeline/) and the publication checklist
(quality_control/publication-checklist.md) before delivery.
- No silent failures. Anything missing surfaces in a
Known gaps
block at the end of the paper.
- Multi-agent ready. For long papers, dispatch sub-agents per
orchestration/agents.md.
3. Top-level workflow
intake → plan → lit-review → methodology → data-analysis →
visualization → drafting → citations → validation → review → ship
Each step has a dedicated playbook. Read it, do the step, persist the
artifact to disk, move on. Detailed master pipeline:
orchestration/pipeline.md.
4. Format selection
When the user does not specify a format, infer it:
| Signal | Use template |
|---|
| ML / NLP / AI / preprint / arxiv-style | templates/arxiv-paper.md |
| Engineering / hardware / signal / IEEE conference | templates/ieee-paper.md |
| HCI / systems / SIGCHI / SIGGRAPH / ACM | templates/acm-paper.md |
| Biology / medicine / Nature / Science / structured | templates/nature-paper.md |
| Social science / business / humanities / Harvard | templates/harvard-paper.md |
| Literature / systematic / scoping / meta review | templates/literature-review.md |
| Thesis chapter / dissertation | templates/thesis-chapter.md |
| Whitepaper / industry / enterprise | templates/whitepaper.md |
| Survey / state-of-the-art | templates/survey-paper.md |
| Policy brief / regulatory | templates/policy-paper.md |
If still ambiguous, ask once, then proceed.
5. Citation style selection
Map domain → default style if not specified:
- CS / engineering / physics → IEEE numeric
- ML / AI / preprint → author–year (Harvard / APA-compatible)
- Biology / medicine / Nature → Nature numeric superscript
- Social science / business / humanities → Harvard (or APA)
- Law / history → Chicago
Style rules: citation_engine/citation-styles.md. Per-style modules:
citation_engine/styles/. The deterministic formatter is
toolchains/format_bibliography.py.
6. Visualization decision (summary)
Full rules: visualization_engine/decision-engine.md and
visualization_engine/visualization-guide.md. Rendering happens via
toolchains/generate_charts.py; if Python is unavailable, the skill
falls back to Markdown tables + Mermaid diagrams — never silently
skips a planned figure.
| Communication goal | Recommended figure |
|---|
| Compare discrete categories | Bar / horizontal bar / lollipop |
| Show trend over time | Line / multi-line |
| Show distribution | Histogram / violin / box plot |
| Show relationship | Scatter + regression line |
| Show correlation among many vars | Heatmap |
| Show parts of a whole | Stacked bar (preferred over pie) |
| Show flow / transformation | Sankey |
| Show structure / pipeline | Architecture / flowchart |
| Show process / decision | Mermaid flowchart |
| Show geography | Choropleth / point map |
| Show timeline of events | Timeline / Gantt |
| Show conceptual hierarchy | Mind map / tree |
| Side-by-side metrics | Comparative table |
7. Tooling expectations
This skill works in three tiers, gracefully degrading:
| Tier | Capabilities |
|---|
| 0. Pure prose (no tools) | Outline + draft + Markdown tables + Mermaid diagrams |
| 1. + Filesystem read/write | Persist sections, bibliography, validation reports |
| 2. + Python (pandas/matplotlib) | Real charts (PNG + SVG), statistical validation, data analysis |
| 2+. + Web search / fetch | DOI verification, source retrieval, retraction checks |
| 2+. + Pandoc (optional) | Output to PDF / DOCX / HTML / LaTeX / RTF / EPUB / ODT / PPTX |
If a tier is missing, the skill detects it and adapts — no silent failures.
See toolchains/README.md for setup.
Output formats
The skill produces Markdown by default. For other formats, run the
output converter:
python toolchains/convert_output.py --input paper-final.md --to pdf --out paper.pdf
python toolchains/convert_output.py --input paper-final.md --to docx
python toolchains/convert_output.py --input paper-final.md --to html
python toolchains/convert_output.py --input paper-final.md --to tex
python toolchains/convert_output.py --input paper-final.md --to epub
Supported targets (via Pandoc): md (always), html, docx, pdf
(needs LaTeX), tex, rtf, epub, odt, pptx.
Self-test:
python toolchains/convert_output.py --self-test
The user can also request a non-Markdown output directly:
/research "topic" --output paper.pdf
/research "topic" --output paper.docx
8. Output contract
Every artifact this skill produces includes, at minimum:
- Title — specific, ≤ 15 words.
- Authors / Affiliation block — placeholders if not provided.
- Abstract — 150–300 words, structured.
- Keywords — 4–8.
- Plain-English summary — 5–10 sentences.
- Numbered sections following the chosen template.
- At least one figure and one table for any paper > 1500 words
(unless purely theoretical and explicitly opted out).
- In-text citations in the chosen style.
- Full reference list with DOIs / URLs.
- Limitations section.
- Future work section.
- Reproducibility statement (data, code, environment, seeds).
- Appendices for derivations, hyperparameters, prompts, raw outputs.
Anything missing is surfaced in a final Known gaps block —
never silently swallowed.
9. Long-context strategy
For papers > ~10,000 words:
- Persist every artifact to disk before moving on
(
paper-spec.md → outline.md → bibliography.yaml →
methodology.md → analysis/findings.md → figures-plan.md →
sections/<NN>-<name>.md → paper-draft.md → paper-cited.md →
paper-final.md).
- Read only the section being drafted (plus the outline) at any time.
- Cross-section consistency is enforced by the outline + a final
cover-to-cover read pass.
Full strategy: long_context/strategy.md.
10. Multi-agent orchestration
For deep / parallel runs, dispatch sub-agents:
| Agent | Reads | Writes |
|---|
| Researcher | prompts/literature-search.md | bibliography.yaml, lit-themes.md |
| Methodologist | prompts/methodology-design.md | methodology.md |
| Analyst | prompts/data-analysis.md | analysis/findings.md |
| Visualizer | prompts/visualization-planning.md | figures-plan.md, figures/ |
| Writer (×N) | prompts/writing-prompts.md | sections/<NN>-<name>.md |
| Citator | prompts/citation-prompts.md | paper-cited.md |
| Validator | validators/ | validation/ |
| Reviewer (×3) | prompts/review-prompts.md | review/ |
Topology: orchestration/agents.md.
11. Failure handling
- Missing data → synthetic illustrative dataset, clearly labeled.
- Unverifiable source →
[UNVERIFIED], listed in Known gaps.
- Conflicting evidence → explicit "Contradictions in the literature"
subsection.
- Out-of-scope request → narrow scope, list dropped sub-topics in
Future work.
- Token / context pressure → see §9.
Full failure-handling matrix: orchestration/failure-handling.md.
12. Where to look next
- Plan a paper →
orchestration/pipeline.md
- Pick a template →
templates/
- Write a section →
prompts/writing-prompts.md
- Add citations →
citation_engine/, workflows/citation-pipeline.md
- Make charts →
visualization_engine/, workflows/visual-generation-pipeline.md
- Validate stats →
methodology_engine/statistical-methods.md,
toolchains/statistical_validation.py
- Self-review →
review_pipeline/, rubrics/academic-quality.md
- Ship it →
quality_control/publication-checklist.md
Always prefer reading the specific file you need over re-reading this one.