paper-radar-jev
Python CLI that fetches recent arXiv papers, scores relevance to your research profile with Jev P(True) and writes daily rankings.
Overview
paper-radar-jev is a configurable Python CLI that fetches recent arXiv papers, scores each one's relevance to a research profile using TypeSafe's P(True) Noul answer, and generates daily Markdown and JSON rankings. The default profile focuses on computer vision and generative models, and you edit config/config.json to set your own interests.
It sends your profile, paper titles, abstracts and categories to TypeSafe for evaluation; reports go to daily/md and daily/json, and history and processed IDs are stored so papers are not re-scored. A check-config option validates configuration without calling the API, and the README suggests a three-paper smoke test first. It requires Python 3.11+, uv, Git, GitHub CLI and your own TypeSafe key. It is MIT licensed with a Traditional Chinese README.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Yes / no | Is this paper relevant to my research profile? | true · false |
How it works
- Fetch recent arXiv papers per configuration.
- Ask Jev P(True) for relevance to the profile.
- Rank papers by probability.
- Write daily Markdown and JSON reports.
Strengths
- Simple configurable daily ranking.
- Skips already-processed papers.
- Config validation before spending.
Caveats
- Requires a TypeSafe key.
- Ranking only; no published feed.
Getting started
gh repo clone LYchoon/paper-radar-jev && cd paper-radar-jev && uv sync, copy .env and config examples, then uv run paper-radar --config config/config.json.At a glance
- Type
- Project
- Status
- prototype
- Language
- Python
- Runtime
- CLI
- Has code
- yes
- License
- MIT
Use cases and integrations
How curated lists describe it
An automated research paper radar that fetches the latest papers from arXiv, evaluates their relevance to a configurable research profile using TypeSafe AI, and ranks them by relevance score. Designed for personalized, daily literature discovery across different research domains.
logicrw/awesome-jev-projects
Frequently asked questions
What is sent to TypeSafe?
Your research profile and each paper's title, abstract and categories.
Where are results saved?
In daily/md and daily/json, with history in data/.
How is relevance scored?
By Jev's probability that the paper is relevant (P(True)).
Related projects
Local-first feed that merges arXiv, GitHub, Hacker News and Bilibili results, with optional Jev title-relevance scores as a filter.
ProjectbetaTypeScriptChrome extension that scores loaded X posts against your reading goal with Jev and ranks top matches pairwise.
ProjectbetaJavaScriptTech reading feed from Hacker News, Dev.to, Hashnode and Lobsters, ranked by Jev scores against your skill summary.
ProjectprototypeTypeScriptChrome extension that asks Jev typed questions about an article, video or feed item and says read now, skim, save or skip.
Browser extensionbetaChecks whether each cited paper supports its citing sentence: code fetches papers, Claude proves a quote, Jev scores it, a human decides.
ServiceproductionReal-time global intelligence dashboard; per the lists, it batches Jev questions to score headline severity and classify event domains.
AppproductionTypeScript