Turn an open-source news engine into an industry daily brief

Share

The AI news site AIHOT has open-sourced its engine under the MIT license. This covers what the stack includes, what it leaves out, the costs to watch, and a prompt that has your AI agent adapt it to your own industry.

Machine translation. Read the original

SourceGitHubReported Nov 19, 2026Not yet tested

If you want a website that watches your industry every day and turns it into a daily briefing, there's now an open-source framework you can start from. AIHOT, an AI news site, has put its whole engine on GitHub under the MIT license, so you can use it commercially. Swap in sources and a definition of "news worth reading" for your field, such as law, finance, or local news, and it becomes your own industry news site.

What it is

AIHOT (aihot.news) is a Chinese-language AI news site. According to the official repo, it collects material from a set of sources every day. A large model filters it first, then scores each item twice on its own. The items worth reading get a Chinese headline and summary. Stories about the same event from different sources get grouped together, ranked by how much they're discussed, and published as a daily briefing each morning.

What's open source is the engine: the whole site and admin panel, the selection and scoring pipeline, and the clustering and ranking logic. The full prompts and the selection criteria are in the repo too.

Item Details (from the official repo)
License MIT
Stack TypeScript, Node.js 24, PostgreSQL 17
Deployment Docker Compose
Sample sources 18 public English AI news sources, enough to see it run
Entry points for agents MCP and the /agent page on the site

Not included: AIHOT's own source list, its operating data, and AI-only features like the model leaderboard. The author also asks that you don't use the AIHOT name or logo.

What it means for you

You know your industry. You know which websites, newsletters, and regulatory notices are worth reading. That's exactly the part the engine doesn't have. The repo includes a guide to "turning it into your own industry" that mainly covers swapping sources and changing the selection criteria. Some users have already adapted it into a local news briefing and their own news Skill. That's user feedback, and the results aren't verified yet.

Hand it to your AI agent to adapt

I want to use the open-source project AIHOT (https://github.com/KKKKhazix/AIHOT) to build a daily news briefing site for [your industry, e.g. cross-border e-commerce]. Please:
1. Read the repo's README and docs/customize.md first. In plain language, tell me what I need to prepare (a server, a database, a model API key) and roughly how much it costs per month.
2. List 15 sources worth tracking in this industry, and mark which ones have an RSS feed and which need web scraping.
3. Rewrite the "what counts as news" selection criteria for my industry. Show me the before and after first.
4. Run it locally with Docker Compose. First confirm it produces a daily briefing using only the sample sources, then switch to my own sources.
5. Stop and ask me before installing any software, entering any keys, or paying for anything.

Things to watch

  • Collecting sources and scoring them both call a large model every day, so costs rise as you add sources. Start with a few sources and check the bill for a week.
  • Before scraping a website, read its terms of use. Don't copy other people's full articles onto your own site.
  • The author says the open-source version isn't guaranteed to match the live site's latest updates. If you run into a problem, check the repo's Issues and Discussions first.
  • To compare other automated briefing setups, see Horizon watches the tech world for you and writes a daily briefing.

References

Done with this one? Mark it to track your path.Sign in
Save it for laterSign in
Questions about this piece?Ask AI about this

Comments (0)

No comments yet. Start the conversation.

Sign in to join the discussion.Sign in