skillfeed
Tech reading feed from Hacker News, Dev.to, Hashnode and Lobsters, ranked by Jev scores against your skill summary.
Overview
Skillfeed pulls the day's writing from Hacker News, Dev.to, Hashnode and Lobsters and ranks it against a short summary of what you know and care about. You pick the platforms, describe your skills once (plus an optional avoid list), and get the best matches first.
Jev receives your skill summary and a batch of article metadata (title, description and tags, without scraping full pages) as state, with one Score question per article on a poor-to-excellent rubric for how well it matches your skills. It is called through the Vercel AI SDK's experimental_evaluate API on AI Gateway, in batches of seven by default to stay under the 32K context window, and the feed is sorted by the returned scores. A provider adapter allows swapping evaluation backends, and a server-sent-events UI shows progress. It is a Next.js app deployed on Vercel.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Score | How well does this article match the user's skills? | poor · excellent |
How it works
- Fetch metadata from selected platforms in parallel
- Batch articles with the user's skill summary as state
- Ask Jev one Score question per article
- Sort the feed by score
Strengths
- Cheaper than generating text per article
- Batches stay within context limits
- Swappable evaluation backend
Caveats
- Needs a Vercel AI Gateway key
- Ranks on metadata only, not full text
Getting started
Clone, npm install, set a Vercel AI Gateway key, then run the dev server; or use the live demo.At a glance
- Type
- Project
- Status
- prototype
- Language
- TypeScript
- Runtime
- Next.js on Vercel
- Has code
- yes
- License
- MIT
Use cases and integrations
How curated lists describe it
A tech reading feed ranked to your skills — powered by TypeSafe Jev.
hellogumbo/awesome-jev
Frequently asked questions
What does Jev decide?
How well each article matches your skill summary, on a poor-to-excellent rubric.
How is Jev called?
Through the Vercel AI SDK's experimental_evaluate on AI Gateway.
Does it read full articles?
No, only titles, descriptions and tags.
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.
ProjectbetaJavaScriptChrome extension that asks Jev typed questions about an article, video or feed item and says read now, skim, save or skip.
Browser extensionbetaPython CLI that fetches recent arXiv papers, scores relevance to your research profile with Jev P(True) and writes daily rankings.
ProjectprototypePython- Paper Radar★ 30
Daily arXiv and bioRxiv radar where Jev answers one yes/no per plain-English interest for every new paper; fork-and-go GitHub Action.
AppbetaPython Multi-user Telegram bot that judges every new post in public channels with Jev templates and routes matches to your DM or channels.
ProjectbetaPython