papertools - Operate the PaperTools arXiv paper reading pipeline
Operates the PaperTools arXiv reading pipeline in Agent-native or API-backed mode to crawl, filter, summarize, browse, and publish paper reading results.
Tags
Updated: 2026-09-22Capabilities
Typical Inputs
What this skill does
- Crawl arXiv paper metadata
- Filter papers using research rubric
- Summarize selected papers
- Generate daily reading results
- Serve published web pages
- Customize paper filtering scope
- Validate publication payload quality
- Resume failed pipeline stages
Inputs
- Target date
- arXiv categories
- Research interests
- API keys and provider credentials
Outputs
- Crawled arXiv paper metadata files
- Agent mode output files
- Published webpage payload files
- Local HTTP server preview endpoint
Requirements
- Python 3.10 or newer
- PaperTools repository root workspace
- Model API credentials for full pipeline
