learnprompt/ai-news-radar

24-Hour AI Radar Update | Sanwei Sharp Review

**First, I’ll help you pick out the best news from a bunch of sources, then merge scattered news into a storyline, and finally sharply review the daily headlines with three different flavors. **

After installing, he said to the Agent: 今天AI圈有什么？

Online Site → news.learnprompt.pro (Data Source/Backup: learnprompt.github.io/ai-news-radar)

English · Radar Skill · Bole Skill · Information source strategy

Update Notes: v0.9 converged the interface into a single-layer information architecture (tab × sections, Featured/Full, × timeline). Old three-view screenshots are archived at /legacy/, retained until mid-August 2026.

---

Choose a side to get in 30 seconds

(1) Let the Agent Read the Newspaper for You → The above line of the installation command, after installing one sentence, get the presentation, zero API, zero Key, zero server:

(2) View web pages directly → Open news.learnprompt.pro. By default, it’s the mobile version’s view. The “View” switch in the upper right corner can switch to the classic version (old desktop interface, path /classic/), or you can use ?view=mobile / ?view=classic / ?view=auto to specify that both views read the same data/ directory data. From v0.9 onward, it uses a single-layer information architecture: top tab for “All/Model/Product/Developer/Industry/Paper/Community/Self-Media” section + “Featured/Complete” global switch. The main list is grouped in reverse order by time and by day, and the “Current Trend” list does not have a fixed number of entries to view the hottest current item. Each featured card comes with a sentence “Recommended Reason”; When the same event is reported by multiple sources, it is folded into a “Multi-source N” tag, allowing you to view each company’s independent headline.

(3) Fork with your own sieve → Fork this repository, change the source to your own OPML, and modify the markdown file under personas/, with data stored on your own GitHub Pages. Jump to fork guide.

The third layer is one path: let agents read newspapers → read newspapers themselves→ run newspapers themselves.

---

Sister Stations · An assembly line

Radar tube tracking, with one station each upstream and downstream. First, go to AI MAP Hotspot Map to see who has been heating up in the past 7 days (refreshing every 12 hours). After locking onto the target, return to the radar to dig deeper into the source; When you see a hotspot you want to get started, go to the GoodCase use case library to directly check existing cases and prompts. Discover → Track → usage, three stops, one path.

---

What is this?

AI News Radar is an automatically updated 24-hour AI update radar. It doesn’t just capture AI news; it first assesses the quality of information sources, merges the same event into a storyline, then scores and reviews you with three types of personas, and finally uses Bole Select, AI tags, source health, and AI proportions to help you judge:

What information is worth watching, what is worth digging into, and what is just noise.