v2.2.0 - Clinical Reports Skill & Enhanced Medical Documentation

🎉 What's New in v2.2.0

✨ New Capabilities: Clinical Reports Skill

This release introduces comprehensive clinical documentation capabilities, expanding Scientific Writer beyond academic papers into the medical and healthcare domain.

🏥 Clinical Reports System

The new Clinical Reports Skill provides a complete clinical documentation framework:

Four Major Report Types:

• ✅ Case Reports - CARE-compliant case report writing for journal publication
• ✅ Diagnostic Reports - Radiology (ACR), Pathology (CAP), Laboratory (CLSI) standards
• ✅ Clinical Trial Reports - SAE reports, CSRs following ICH-E3 guidelines
• ✅ Patient Documentation - SOAP notes, H&P, discharge summaries, consultations

Professional Templates & Standards:

• 📋 12 industry-standard templates for immediate use
• 📚 8 comprehensive reference guides (570-745 lines each)
• 🐍 8 Python validation and automation scripts
• ✅ Quality assurance checklists

Compliance & Standards:

• 🔒 HIPAA compliance and de-identification tools
• 📜 Regulatory compliance (FDA 21 CFR Part 11, ICH-GCP)
• 🏷️ Medical terminology standards (SNOMED-CT, LOINC, ICD-10, CPT)
• 🔗 Integration with scientific-writing and peer-review skills

🔧 Improvements

• Enhanced medical and clinical documentation capabilities
• Expanded document generation beyond academic papers to clinical settings
• Added healthcare regulatory compliance features
• Improved template library with industry-standard medical formats

📝 Documentation Updates

• Updated README.md to include clinical reports in document generation
• Updated docs/SKILLS.md with comprehensive clinical-reports skill documentation
• Updated docs/FEATURES.md with clinical reports examples
• Added clinical-reports/README.md with quick start guide

📦 Installation

pip install scientific-writer==2.2.0
# or
uv pip install scientific-writer==2.2.0
# or
uvx scientific-writer@2.2.0

🚀 Quick Start - Clinical Reports

scientific-writer
> Generate a case report for a 45-year-old patient with acute MI following CARE guidelines

import asyncio
from scientific_writer import generate_paper

async def main():
    async for update in generate_paper(
        "Create a CARE-compliant case report for acute myocardial infarction"
    ):
        if update["type"] == "result":
            print(f"Report: {update['files']['pdf_final']}")

asyncio.run(main())

---

Full Changelog: https://github.com/K-Dense-AI/claude-scientific-writer/blob/main/CHANGELOG.md

v2.1.0 - Citation Management & Venue Templates

🎉 What's New in v2.1.0

✨ New Skills

Citation Management Skill

Advanced citation quality control system that helps ensure your papers meet publication standards:

• Validates all citation metadata for completeness and accuracy
• Checks for proper author names, titles, venues, DOIs, and URLs
• Reduces AI hallucinations in bibliographic references
• Helps avoid citation-related desk rejects

Venue Templates Skill

Comprehensive academic submission templates for streamlined publishing:

• Journal templates: Nature, Science, Cell, PNAS, and more
• Conference templates: NeurIPS, ICML, CVPR, ACL, and more
• Poster templates with venue-specific dimensions and styles
• Grant proposal templates: NSF, NIH, DOE, DARPA
• Venue-specific formatting guidelines and requirements
• Reference documents with submission best practices

🔧 Improvements

• Enhanced citation accuracy through automated metadata validation
• Streamlined academic submission workflow with ready-to-use templates
• Better support for multiple publication venues and formats

📦 Installation

# Using pip
pip install scientific-writer==2.1.0

# Using uv
uv pip install scientific-writer==2.1.0

# Using uv tool (for CLI)
uv tool install scientific-writer

# Using uvx (one-off CLI usage)
uvx scientific-writer

📚 Documentation

• Full documentation: README.md
• API Reference: docs/API.md
• Features Guide: docs/FEATURES.md
• Skills Documentation: docs/SKILLS.md

Full Changelog: https://github.com/K-Dense-AI/claude-scientific-writer/blob/main/CHANGELOG.md

v2.0.1 - Documentation Updates

📝 Documentation Updates (v2.0.0 → v2.0.1)

Note: This is a documentation-only release. All functionality from v2.0.0 is preserved with no code changes, no new features, and no breaking changes.

Version 2.0.1 improves documentation for existing features without changing any code. If you're upgrading from v2.0.0, you only get better docs - the software is functionally identical.

---

What Changed from 2.0.0

New Documentation Files

• docs/FEATURES.md - Comprehensive guide documenting existing features:
  • Document generation types (papers, posters, grants, reviews, schematics)
  • AI-powered capabilities (research lookup, peer review, iterative editing)
  • Intelligent paper detection system
  • Data & file integration workflows
  • Document conversion with MarkItDown
  • Developer features and API patterns

Enhanced Existing Documentation

• README.md - Reorganized structure:

  • Categorized features (Document Generation, AI Capabilities, Developer Tools)
  • Expanded CLI and API usage examples
  • Added workflow examples for common use cases
  • Better visual organization

• docs/API.md - Added advanced topics:

  • Research lookup setup and usage patterns
  • Data file processing workflows
  • Intelligent paper detection explanation
  • Custom output organization patterns
  • Metadata extraction examples
  • Progress monitoring patterns (progress bars, stage-based, logging)
  • Multiple paper generation (sequential and parallel)

• Documentation structure - Reorganized into:

  • User Guides: Features, API, Skills, Troubleshooting
  • Developer Resources: Development, Releasing, Changelog, System Instructions

Features That Were Already There (Now Better Documented)

These features existed in v2.0.0 and earlier but lacked comprehensive documentation:

• 📊 Figure generation & visualization - Python-based scientific schematics (matplotlib, graphviz, schemdraw, networkx)
• 🔬 Scientific diagrams - CONSORT flowcharts, circuit diagrams, biological pathways, neural network architectures
• 🔍 Research lookup - Real-time literature search with Perplexity Sonar Pro via OpenRouter
• 🎯 Intelligent paper detection - Automatic context tracking in CLI
• 📝 Grant proposals - NSF, NIH, DOE, DARPA with agency-specific formatting
• 🔄 Document conversion - 15+ formats with MarkItDown (PDF, DOCX, PPTX → Markdown)
• 📈 ScholarEval framework - 8-dimension quantitative paper evaluation
• 🎓 Research posters - LaTeX-based conference posters (beamerposter, tikzposter, baposter)

Important: None of these are new in 2.0.1 - we just documented them properly!

---

What's in v2.0.0 (The Previous Release)

v2.0.0 was the major release that added the programmatic Python API:

• ✨ New: Programmatic Python API (generate_paper() function)
• ✨ New: Real-time progress updates via async generators
• ✨ New: Type hints and data models (ProgressUpdate, PaperResult, etc.)
• ✨ New: Modular package structure (api.py, cli.py, core.py, models.py, utils.py)
• ✅ Preserved: 100% CLI backward compatibility - all v1.x CLI commands work identically

See the v2.0.0 release notes for details on the API addition.

---

Installation

# Using pip
pip install scientific-writer==2.0.1

# Using uv (recommended)
uv pip install scientific-writer==2.0.1

# Run directly with uvx
uvx scientific-writer@2.0.1

Quick Start

CLI:

scientific-writer
> Create a Nature paper on CRISPR gene editing

Python API:

import asyncio
from scientific_writer import generate_paper

async def main():
    async for update in generate_paper("Create a Nature paper on CRISPR"):
        if update["type"] == "progress":
            print(f"[{update['percentage']}%] {update['message']}")
        else:
            print(f"PDF: {update['files']['pdf_final']}")

asyncio.run(main())

Generate figures programmatically:

# Figures are generated automatically during paper creation
# Or request them explicitly:
async for update in generate_paper(
    "Create a transformer architecture diagram for my NeurIPS paper"
):
    pass  # Diagram saved to figures/ directory

---

Full Changelog: https://github.com/K-Dense-AI/claude-scientific-writer/blob/main/CHANGELOG.md#201---2025-10-30

v1.1.1 - Troubleshooting Guide

What's New in v1.1.1

Documentation Improvements

• Added comprehensive Troubleshooting Guide covering:
  • Installation and dependency issues
  • API key configuration problems
  • LaTeX compilation errors
  • Common workflow issues
  • Debugging strategies

This release enhances the user experience by providing detailed guidance for resolving common issues encountered when using the Claude Scientific Writer.

Full Changelog: v1.1.0...v1.1.1

🚀 Release v1.1.0 - Enhanced Intelligence & Workflow

We're excited to announce version 1.1.0 of Claude Scientific Writer, packed with powerful new features and improvements to streamline your scientific writing workflow!

✨ Major Features

🧠 Intelligent Paper Detection

• Automatic paper recognition: The CLI now automatically detects when you're referring to a previous paper
• Smart context switching: Just use keywords like "continue", "update", or "the paper" to resume work
• Topic-based detection: Reference papers by topic (e.g., "fix the acoustics paper")
• Session continuity: Seamlessly switch between multiple papers in a single session

📦 Modern Package Management

• UV integration: Migrated from requirements.txt to pyproject.toml with uv package manager
• Faster installations: Improved dependency resolution and installation speed
• Better compatibility: Enhanced cross-platform support and reproducibility

📄 Enhanced Document Processing

• MarkItDown + OpenRouter: AI-powered document conversion with advanced vision models
• Image descriptions: Automatic, detailed descriptions for images in documents
• Expanded format support: Enhanced handling of DOCX, PDF, PPTX, XLSX files
• Smart conversions: Context-aware document transformations

🔬 Intelligent Research Lookup

• Automatic model selection: Queries are routed to appropriate models based on complexity
• Sonar Reasoning Pro: Advanced reasoning for complex research questions
• Better error handling: Improved response logging and user feedback
• Real-time integration: Seamless Perplexity Sonar Pro integration

🎯 New Capabilities

📊 Data File Integration

• Automatic processing: Place files in data/ folder for automatic inclusion
• Smart organization: Images → figures/, data → data/ subdirectories
• Seamless workflow: Files automatically incorporated into your papers

📝 LaTeX Research Posters

• Professional templates: Publication-ready poster layouts
• Design principles: Built-in best practices for academic posters
• Quality checks: Automated validation scripts for poster formatting

✅ Quality Assurance

• PDF formatting reviews: Comprehensive visual inspection protocols
• Citation verification: Strict verification ensuring all citations are real and verifiable
• Iterative improvements: Automated quality check workflows
• Research integrity: Enhanced logging and metadata verification

🌐 Additional Tools

• paper2web: Convert papers to web-friendly formats
• Peer review system: Comprehensive evaluation stages and report generation
• Smart automation: Improved workflow automation throughout the pipeline

🛠️ Improvements

• Enhanced README documentation with clearer examples and workflows
• Refined new paper detection logic for better accuracy
• Improved .gitignore for better file management
• Better user guidance for data file preparation
• Enhanced error messages and user feedback

📚 Documentation Updates

• Comprehensive installation instructions
• API key setup guides for OpenRouter integration
• Detailed workflow examples
• Quality assurance guidelines
• Research lookup best practices

🔄 Migration Notes

If upgrading from v1.0.0:

1. Install uv package manager: curl -LsSf https://astral.sh/uv/install.sh | sh
2. Run uv sync to update dependencies
3. Review the updated README for new features

🙏 Thank You

Thank you to everyone who provided feedback and helped shape this release!

v0.1 - Initial Release

Claude Scientific Writer v0.1

🎉 Initial Release

A powerful command-line tool for scientific writing powered by Claude Sonnet 4.5 and the Claude Agents SDK.

✨ Features

Core Capabilities

• 🔬 Scientific Writing: Generate complete scientific papers following IMRaD structure with proper citations and formatting
• 📚 Literature Review: Conduct comprehensive literature reviews with citation management
• 🔍 Research Lookup: Real-time research information using Perplexity's Sonar Pro model via OpenRouter
• 📝 Peer Review: Receive constructive feedback on scientific drafts
• 🤖 LaTeX Support: Generates publication-ready LaTeX documents with BibTeX citations

Document Manipulation

• 📄 Work with various document formats (docx, pdf, pptx, xlsx)
• 📊 Progress tracking with real-time logging
• 📦 Automatic data file and image integration

Data File Integration

• Automatically process and incorporate data files from the data/ folder
• Smart routing of images to figures/ and data files to data/
• Support for CSV, JSON, Excel, PNG, JPG, SVG, and more

🚀 Quick Start

Prerequisites

• Python 3.10 or later
• An Anthropic API key
• An OpenRouter API key (for research lookup)

Installation

git clone https://github.com/K-Dense-AI/claude-scientific-writer.git
cd claude-scientific-writer
pip install -r requirements.txt

Setup API Keys

echo "ANTHROPIC_API_KEY=your_key_here" > .env
echo "OPENROUTER_API_KEY=your_key_here" >> .env

Usage

python scientific_writer.py

📖 Documentation

See README.md for complete documentation and SKILLS.md for detailed skill documentation.

🙏 Acknowledgments

• Built with the Claude Agents SDK
• Powered by Anthropic's Claude Sonnet 4.5
• Research lookup powered by Perplexity Sonar Pro via OpenRouter

📄 License

MIT License - see LICENSE file for details.