This project provides a customizable, AI-powered daily news brief that aggregates content from over 26 diverse data sources, including tech news, AI advancements, global finance, and world politics. It targets technically savvy users who desire a self-hosted, consolidated digest of information, enhanced with LLM-generated summaries and market analysis, all controllable on their own infrastructure.

How It Works

The system aggregates data from numerous sources (RSS, APIs, scraping) and processes it using configurable LLM backends (Anthropic, OpenAI, DeepSeek, etc.) for summarization and market commentary. It offers flexible deployment options, including GitHub Actions for zero infrastructure, local OS scheduling, or self-hosted servers, outputting a single, self-contained HTML file.

Quick Start & Requirements

• Installation: Three primary methods:
  1. GitHub Actions + Pages: Fork the repository, configure secrets/variables. Estimated setup: ~5 minutes.
  2. Local One-liner: Execute a curl or PowerShell command piped to node. Estimated setup: ~3 minutes.
  3. AI Agent: Instruct an AI agent (e.g., Claude Code, Cursor) to install via the README.
• Prerequisites: Node.js 20+, npm, git (for local installs). A working LLM backend (Claude Code CLI logged in, or API key for supported providers).
• Platform: Windows 10/11, macOS 12+, Linux.
• Links: Project Repository, Live Demo A, Live Demo B.

Highlighted Details

• Multi-source Aggregation: Default enablement of 26+ sources across tech, AI, finance, world news, and community discussions.
• Market Data Analysis: Integrates real-time data for 21 tickers (US stocks, crypto, HK stocks, commodities, macro signals) with technical indicators and AI-driven trading commentary.
• Pluggable LLM Backends: Supports 6 LLM providers (Claude CLI, Anthropic, OpenAI, DeepSeek, MiniMax, Zhipu) switchable via a single environment variable, avoiding vendor lock-in.
• Bilingual Support: REPORT_LOCALE=en toggles sources, prompts, and UI text to English.
• Zero Data-Source API Keys: All news and information feeds utilize free public endpoints, eliminating subscription costs for data acquisition.
• Single-File HTML Output: Generates an HTML file with inlined CSS and JS, suitable for direct deployment as a static webpage.

Maintenance & Community

The project is shared on the LINUX DO open-source community, but specific details on active contributors, sponsorships, or dedicated community channels (like Discord/Slack) are not explicitly mentioned in the README.

Licensing & Compatibility

The project is released under the MIT License, which permits commercial use and integration with closed-source projects.

Limitations & Caveats

For free-tier GitHub accounts, GitHub Pages requires the repository to be public. macOS and Linux schedulers may not wake sleeping computers, potentially causing missed scheduled runs, unlike Windows Task Scheduler. While data sources are free, LLM API calls incur costs, estimated at under $2 per month for common configurations. Advanced LLM backend configurations, especially with proxies, require careful management of repository secrets and variables.