Claude Scientific Writer
ResearchDeep-research plugin generating publication-ready papers, grants, posters, and clinical reports with verified citations and AI diagrams.
Install
$ /plugin marketplace add https://github.com/K-Dense-AI/claude-scientific-writer
Update
$ /plugin marketplace update claude-scientific-writer
README documents no explicit update command; the reliable path is re-running the marketplace add against the GitHub repo, then restarting Claude Code. PyPI CLI users update via pip install -U scientific-writer.
When to use it
- You need a publication-ready paper, NSF/NIH grant, conference poster, or clinical report drafted from your own CSV/figure data.
- You want every claim backed by real-time literature search and verified citations rather than hallucinated references.
- You need to generate scientific diagrams (CONSORT flowcharts, neural networks, pathways) from natural-language descriptions.
- You want manuscript peer-review feedback scored against the structured 8-dimension ScholarEval rubric.
Commands & flags
- /plugin install claude-scientific-writer
- Install the plugin after adding the K-Dense-AI GitHub marketplace.
- /claude-scientific-writer:scientific-writer-init
- Write a CLAUDE.md in your project and enable all 20+ scientific writing skills.
- @research-lookup
- Real-time literature search via Perplexity Sonar Pro to back claims with verifiable sources.
- @peer-review
- Systematic manuscript evaluation using the 8-dimension ScholarEval framework.
- @scientific-schematics
- Generate CONSORT diagrams, neural architectures, and pathways via Nano Banana Pro.
- pip install scientific-writer
- Install the PyPI package for the standalone scientific-writer CLI and async Python API.
How to use it
Add the marketplace from the GitHub repo, install, restart, then initialize per project:
/plugin marketplace add https://github.com/K-Dense-AI/claude-scientific-writer
/plugin install claude-scientific-writer
# restart Claude Code, then in your project:
/claude-scientific-writer:scientific-writer-init
Skills are invoked with an @ prefix, e.g. @research-lookup Find papers on mRNA vaccine efficacy or @peer-review Evaluate this manuscript. Drop figures and data into your project, then prompt with explicit file references:
> Create a Nature paper on CRISPR gene editing. Present editing_efficiency.csv
(5 cell lines), include Western_blot.png and flow_cytometry.png showing 87%
efficiency (p<0.001). Compare with literature benchmarks.
Set ANTHROPIC_API_KEY (required) and OPENROUTER_API_KEY (optional, for research lookup). For CLI/API use without Claude Code, pip install scientific-writer exposes the scientific-writer command and an async Python API (Python 3.10-3.12).
Why it earns its place
Built by K-Dense Inc as the open-source tier of their commercial scientific-AI platform, it bundles 20+ purpose-built skills (papers, grants, posters, clinical reports, peer review) behind one plugin. Its differentiator is research-first generation: it searches real literature via Perplexity Sonar Pro and grounds every claim in verifiable citations, then handles BibTeX, LaTeX, and figure integration end to end, including AI-generated CONSORT/pathway diagrams via Nano Banana Pro. The same SKILL.md files double as portable prompts for any agentic LLM.
Try it
> Write an NSF grant proposal on quantum error correction. Present gate_fidelity.csv (99.2% vs 95% state-of-art), include circuit_topology.png and error_rates.csv, and add a 5-year timeline from milestones_budget.xlsx.