Academic Research Plugin for Claude Code

A suite of 8 AI-powered academic workflow skills for Claude Code, designed for AI/ML PhD researchers.

Installation

Quick Install (Recommended)

# Add the marketplace
claude plugin marketplace add JeanDiable/academic-research-plugin

# Install the plugin
claude plugin install academic-research

Manual Install

1. Clone this repository:

git clone https://github.com/JeanDiable/academic-research-plugin.git

2. Install Python dependencies:

pip install -r academic-research-plugin/lib/requirements.txt

3. Install the plugin in Claude Code:

claude plugin add /path/to/academic-research-plugin

Optional

Set a Semantic Scholar API key for higher rate limits:

export S2_API_KEY="your-key-here"

Skills

1. Literature Survey (/academic-research:literature-survey)

Comprehensive topic-driven literature survey. Searches arXiv, Semantic Scholar, and DBLP, identifies research gaps, performs cross-domain exploration, and proposes 2-3 innovation directions.

/academic-research:literature-survey "diffusion models for 3D generation"

2. Paper-Triggered Survey (/academic-research:paper-triggered-survey)

Anchored survey starting from a specific paper. Accepts PDF files, arXiv URLs, tweets, or paper titles. Maps the paper's research landscape and proposes extensions.

/academic-research:paper-triggered-survey https://arxiv.org/abs/2301.12345
/academic-research:paper-triggered-survey /path/to/paper.pdf

3. Survey Writing (/academic-research:survey-writing)

Taxonomy-focused survey paper drafting. Searches for papers, builds a multi-level taxonomy, and drafts a complete survey with introduction, per-category deep dives, comparison tables, and future directions.

/academic-research:survey-writing "vision-language models"

4. Paper Reviewing (/academic-research:paper-reviewing)

Conference-style peer review. Supports NeurIPS, ICML, CVPR, ACL, AAAI, ICCV, ICLR formats with adjustable severity. Generates structured reviews with scores, strengths, weaknesses, and questions.

/academic-research:paper-reviewing /path/to/submission.pdf
/academic-research:paper-reviewing /path/to/paper.pdf --venue neurips --severity balanced

5. Paper Polishing (/academic-research:paper-polishing)

Comprehensive draft feedback in ICML meta-review style. Analyzes correctness, motivation, methodology, presentation, visualizations, and citations. Outputs prioritized revision checklist.

/academic-research:paper-polishing /path/to/draft.pdf

6. Citation Assistant (/academic-research:citation-assistant)

Automatic citation insertion for LaTeX manuscripts. Finds uncited claims, searches for correct papers, fetches official BibTeX (never fabricates), and inserts \cite{} commands.

/academic-research:citation-assistant /path/to/paper.tex

7. Homework Machine (/academic-research:homework-machine)

End-to-end assignment completion. Analyzes requirements, conducts research, writes code with tests, drafts academic reports with citations, and performs paraphrasing via translation round-trip (macOS).

/academic-research:homework-machine /path/to/assignment.pdf

8. Poster & Slides Maker (/academic-research:poster-slides-maker)

Academic presentation creator. Generates conference-quality slides (reveal.js HTML) and/or posters (single-page HTML) from a paper PDF or LaTeX project. Standalone files viewable in any browser.

/academic-research:poster-slides-maker /path/to/paper.pdf --format slides --pages 15
/academic-research:poster-slides-maker /path/to/paper.pdf --format poster

Project Structure

academic-research-plugin/
├── .claude-plugin/
│   └── plugin.json              # Plugin manifest
├── lib/                          # Source of truth for shared utilities
│   ├── paper_search.py          # Unified search: arXiv + Semantic Scholar + DBLP
│   ├── bibtex_utils.py          # BibTeX fetch/merge/validate
│   ├── translate_roundtrip.py   # EN→ZH→EN paraphrasing (macOS)
│   └── requirements.txt         # Python dependencies
├── sync.sh                       # Copies lib/ to each skill's scripts/
├── skills/
│   ├── literature-survey/
│   │   ├── SKILL.md
│   │   └── scripts/             # Synced from lib/
│   ├── paper-triggered-survey/
│   │   ├── SKILL.md
│   │   └── scripts/
│   ├── survey-writing/
│   │   ├── SKILL.md
│   │   └── scripts/
│   ├── paper-reviewing/
│   │   ├── SKILL.md
│   │   ├── scripts/
│   │   └── references/
│   │       └── conference_formats.md
│   ├── paper-polishing/
│   │   ├── SKILL.md
│   │   └── scripts/
│   ├── citation-assistant/
│   │   ├── SKILL.md
│   │   └── scripts/
│   ├── homework-machine/
│   │   ├── SKILL.md
│   │   └── scripts/             # Also includes translate_roundtrip.py
│   └── poster-slides-maker/
│       ├── SKILL.md
│       └── assets/
│           └── reveal-template.html
└── README.md

Shared Utilities

paper_search.py

Unified academic paper search across three sources:

• arXiv — Full-text search with metadata
• Semantic Scholar — Citation graph, abstracts, venue info
• DBLP — Conference/journal paper discovery

Features automatic deduplication (DOI exact match + fuzzy title matching at 0.85 threshold), configurable sorting, and JSON/text output.

bibtex_utils.py

BibTeX reference management with a 3-source fallback chain: Semantic Scholar -> DBLP -> arXiv. Never fabricates entries — only fetches from official sources.

translate_roundtrip.py

EN -> ZH -> EN translation round-trip for anti-plagiarism paraphrasing. Uses Google Translate for EN->ZH and Apple's macOS Translation framework for ZH->EN. Preserves technical terms via regex extraction.

Development

After editing any file in lib/, run sync to distribute changes:

./sync.sh

Requirements

• Claude Code
• Python 3.9+
• macOS (for translate_roundtrip.py Apple Translation — other features work cross-platform)
• Optional: S2_API_KEY environment variable for Semantic Scholar API

License

MIT