ArXiv Watcher

Monitor and track new papers on arXiv.org by category.

Overview

This skill helps you:

• Fetch latest papers from specific arXiv categories
• Generate formatted Markdown summaries
• Star/bookmark papers for later reading

Quick Start

# Check latest papers in a category
/arxiv-watcher cs.AI

# Multiple categories
/arxiv-watcher cs.AI cs.CL cs.LG

# View starred papers
/arxiv-watcher --starred

Available Categories

Computer Science (cs.*)

Category	Description
cs.AI	Artificial Intelligence
cs.CL	Computation and Language
cs.CV	Computer Vision
cs.LG	Machine Learning
cs.NE	Neural and Evolutionary Computing
cs.RO	Robotics
cs.SE	Software Engineering

Physics (physics.*)

Category	Description
astro-ph	Astrophysics
cond-mat	Condensed Matter
hep-ex	High Energy Physics - Experiment
hep-th	High Energy Physics - Theory
quant-ph	Quantum Physics

Mathematics (math.*)

Category	Description
math.AG	Algebraic Geometry
math.CO	Combinatorics
math.DG	Differential Geometry
math.NT	Number Theory
math.ST	Statistics Theory

Statistics (stat.*)

Category	Description
stat.ML	Machine Learning
stat.TH	Statistics Theory

Full category list: https://arxiv.org/category_taxonomy

Commands

Fetch Latest Papers

python scripts/arxiv_watcher.py fetch <category> [--limit N]

Fetches recent papers from arXiv. Outputs Markdown format with:

• Paper title and arXiv ID
• Authors
• Abstract preview (first 200 chars)
• Link to full paper

Star Papers

python scripts/arxiv_watcher.py star <arxiv_id>

Adds a paper to starred list for later reference.

Unstar Papers

python scripts/arxiv_watcher.py unstar <arxiv_id>

Removes a paper from starred list.

View Starred Papers

python scripts/arxiv_watcher.py starred

Lists all starred papers with their details.

Output Format

Papers are displayed in Markdown format:

## [2403.12345] Paper Title Here

**Authors:** John Doe, Jane Smith
**Category:** cs.AI
**Submitted:** 2024-03-15

**Abstract:**
This paper presents a novel approach to...

**Links:**
- arXiv: https://arxiv.org/abs/2403.12345
- PDF: https://arxiv.org/pdf/2403.12345.pdf

---

Starred papers include a ⭐ marker for easy identification.

Workflow

When user asks to check arXiv:

1. Identify categories - Parse user request for category names
2. Fetch papers - Run arxiv_watcher.py fetch <category>
3. Format output - Present papers in readable Markdown
4. Offer actions - Ask if user wants to star any papers

Example Interactions

User: "check new papers in cs.AI"

Assistant: Runs fetch command, displays 5-10 recent papers, asks if any should be starred.

User: "star the second one"

Assistant: Parses the arxiv_id from previous output, runs star command, confirms action.

User: "show my starred papers"

Assistant: Runs starred command, displays list with details.

Resources

scripts/

• arxiv_watcher.py - Main CLI for fetching and managing papers

assets/

• starred.json - Persistent storage for starred papers (auto-created)