An open-source framework to aggregate industry news, score items via LLM, and auto-publish daily briefs.
Collect, score, cluster, and publish daily briefings tailored to your specific niche.
AIHOT is a complete website framework that ingests RSS, web pages, JSON APIs, X accounts, and WeChat feeds. It uses an LLM to pre-filter, independently double-score, and summarize articles into a structured industry news station. If you need a domain-specific intelligence feed without building an aggregator from scratch, this is your starting point.
The framework pulls articles, runs duplicate detection and pre-filtering, and applies dual LLM scoring against your custom thresholds to build clean summaries. It clusters related reports from multiple sources into single events, calculates a popularity score based on independent coverage, and generates automated daily, weekly, and monthly digests.
It suits domain experts, researchers, and professionals in fields like law, HR, or finance who want a curated intelligence feed tailored to their specific industry KnowHow. It does not suit non-technical users who expect a one-click SaaS without dealing with Docker, API keys, or prompt adjustments.
Pre-filters and independently scores every raw item twice against your custom entry thresholds.
Groups multiple media reports about the same story into a single event ranked by independent sources.
Captures RSS, web lists, JSON APIs, X accounts, WeChat, and custom script pushes with auto-adjusted frequencies.
Exposes an RSS feed, public API, MCP, and llms.txt so both humans and AI agents can consume the output.
Includes source testing, content diagnostics, selection evaluation tools, and API budget limits.
Free under its MIT license. The only costs are your own server hosting and the API token usage consumed by the LLM when scoring and summarizing incoming articles.
880 open remote AI-training roles, updated November 16, 2026; most pay $20–$175 an hour. Zekai lists them free; the application goes to the employer’s platform.
These reviewed AI tools do the same job instantly — from our scored directory.
Yes. The repository provides 18 AI sources as a demo, but you are meant to replace them and the prompt standards with sources and criteria for your own profession.
It groups reports about the same event together and calculates heat based on unique independent sources over a 48-hour window rather than raw article count.
It runs on Node.js 24, TypeScript, Fastify, PostgreSQL, pg-boss, Tailwind CSS, and React Router with server-side rendering, packaged via Docker Compose.
See Zekai first in Google