Filter each day's new papers down to the few you need
Every morning, Jev reads new arXiv papers against your written interests and keeps a handful, with probabilities you can threshold.
Author's figure: Author's figure: roughly $0.06 per day of papers (Eliot5566). Reported by the builder, not measured by us.
The use case
Researchers and R&D leads cannot keep up with the daily flood of new papers in their field. Here a scheduled job pulls the day's new arXiv listings, and Jev checks each against plain-language interest statements, returning a probability that the paper matters to you. Papers above your threshold go to an email or Slack digest. The reader still opens the papers and decides what is worth time, and should look at the near-misses now and then to tune the interests and threshold.
Example workflow
- Write three to ten interest statements in plain English
- Schedule a daily pull of new arXiv listings with GitHub Actions
- Ask Jev a Noul question per paper per interest
- Send papers above the threshold to a Slack or email digest
- Review near-misses weekly and adjust interests
Tools
- Jev (TypeSafe AI)
- arXiv API
- GitHub Actions
- Slack
Risks and guardrails
- Narrow interests can hide important adjacent work
- Threshold set too high silently drops good papers
- Titles and abstracts can be misleading
Sources
Written by Agents Lookup. Credits and official pages for the tools this use case names:
- JEV-Paper-Radar by Eliot5566 on Made with Jev: madewithjev.com/builds/jev-paper-radar
- JEV-Paper-Radar: original post or repo by Eliot5566: github.com/Eliot5566/JEV-Paper-Radar
- Slack MCP server (official page): docs.slack.dev/ai/slack-mcp-server (checked 2026-11-09)