Paper Pick

Selects one notable research paper daily from Hugging Face Papers and arXiv.

Updated May 25, 2026
One-click install
npx skills add https://github.com/Cbowcrptex/CBOWCRYPTEX --skill paper-pick-cbowcrptex
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: Paper Pick
Source: https://github.com/Cbowcrptex/CBOWCRYPTEX/tree/main/skills/paper-pick
Command: npx skills add https://github.com/Cbowcrptex/CBOWCRYPTEX --skill paper-pick-cbowcrptex

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Keeping up with the flood of new AI research papers is time-consuming, and manually browsing arXiv or Hugging Face every day leads to missed important work or repeated recommendations. ## Core Features & Use Cases - Topic-based search: Queries the Hugging Face Papers API for a given research topic, or browses today's trending daily papers when no topic is set. - Deduplication via memory logs: Reads the last 7 days of memory logs to avoid recommending papers already covered. - Curated single pick: Selects the one most worth-reading paper based on novelty, relevance, practical implications, and community upvotes, then sends a formatted notification and logs the result. - Use Case: A researcher asks for a daily paper on "memory consolidation" and receives one arXiv link with a one-sentence rationale, without re-reading papers suggested earlier in the week. ## Quick Start Ask the assistant to pick today's must-read paper on transformer architectures and send it via the notify script.

Frequently Asked Questions about Paper Pick

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I find trending AI research papers daily?▼

Query the Hugging Face Papers API endpoint /api/daily_papers to get today's trending papers with titles, abstracts, authors, and upvotes. This skill automates that call, picks the single best paper, and sends it via a notify script.

How to search arXiv papers by topic from the command line?▼

Use curl against the Hugging Face Papers search API with a query parameter, for example /api/papers/search?q=your-topic&limit=15. The JSON response includes arXiv IDs, summaries, and upvotes for ranking results.

Does the Hugging Face Papers API require an API key?▼

No, the Hugging Face Papers API endpoints used here are free, require no authentication key, and have no rate limits. Both the search and daily papers endpoints return JSON arrays directly.

How does it avoid recommending the same paper twice?▼

It reads the last 7 days of files under memory/logs/ before selecting a paper and skips anything already mentioned. Each new pick is appended to today's log file for future deduplication.

What happens when curl is blocked in the sandbox?▼

When outbound curl requests are blocked, the skill falls back to the WebFetch tool for any URL retrieval. It can also use WebSearch against arxiv.org to catch papers the API missed.