This project provides an automated workflow for researchers to manage academic papers, leveraging Claude Code skills with arXiv and Semantic Scholar APIs. It targets researchers and power users seeking to streamline the process of discovering, analyzing, and organizing research literature, offering significant time savings and enhanced knowledge synthesis through automated note generation and knowledge graph updates.

How It Works

The system integrates with Claude Code and utilizes APIs from arXiv and Semantic Scholar to automate paper discovery and analysis. Key features include a start-my-day function for daily paper recommendations based on relevance, recency, popularity, and quality, which also generates detailed notes and extracts images for top papers. A paper-analyze function performs deep dives into individual papers, creating structured notes and updating a knowledge graph. Image extraction prioritizes arXiv source files and falls back to PDF extraction, while paper-search allows querying existing notes.

Quick Start & Requirements

• Primary Install/Run: Clone or copy the repository's skill directories (start-my-day, paper-analyze, etc.) into your Claude Code skills directory. Execute commands like start my day or paper-analyze <arXiv_ID>.
• Prerequisites: Claude Code CLI, Python 3.8+, and pip install -r requirements.txt. Requires an Obsidian Vault with a specific directory structure (10_Daily, 20_Research/Papers, 99_System/Config).
• Configuration: Set the OBSIDIAN_VAULT_PATH environment variable. Copy config.example.yaml to config.yaml, edit research interests, and place it in the Obsidian Vault's 99_System/Config/research_interests.yaml.
• Links: QUICKSTART.md (referenced), config.example.yaml, requirements.txt.

Highlighted Details

• Daily Recommendations: Aggregates papers from arXiv (last month) and Semantic Scholar (past year, high-traffic), scoring them on relevance (40%), recency (20%), popularity (30%), and quality (10%).
• Automated Note Generation: Creates structured notes including abstract translation, research background, methods, results, value, limitations, and comparisons.
• Image Extraction: Extracts high-quality images from arXiv source packages or PDFs, saving them into an images/ subdirectory within the paper's note.
• Knowledge Graph Integration: Automatically updates a knowledge graph with analyzed paper information.

Maintenance & Community

The project welcomes Issues and Pull Requests. No specific community channels (e.g., Discord, Slack) or detailed roadmap are mentioned.

Licensing & Compatibility

• License: MIT License.
• Compatibility: The MIT license is permissive, generally allowing for commercial use and integration into closed-source projects.

Limitations & Caveats

Setup requires adherence to specific Claude Code CLI and Obsidian vault configurations. Potential issues may arise from network connectivity, incorrect configuration paths, or inaccuracies in keyword linking, though some aspects are tunable. PDF image extraction is a fallback mechanism and may not always succeed.