OneLine is an AI-driven tool designed for in-depth analysis of news and hot events, transforming unstructured text into structured intelligence reports. It targets researchers, analysts, and power users needing to quickly grasp complex information, offering a five-layer analytical framework to extract facts, sentiments, context, influence, and risks from news articles. The primary benefit is the automated generation of comprehensive, actionable intelligence, saving significant manual analysis time.

How It Works

The core of OneLine is its five-layer analysis framework, powered by AI models from OpenAI or Anthropic. It begins with extracting foundational facts (5W1H), then delves into sentiment and opinion mining, followed by constructing event timelines and identifying correlations. The system further analyzes propagation influence via heat indices and spread paths, culminating in risk assessments and actionable decision recommendations. Integration with the Exa Search API enables intelligent news retrieval, forming the basis for this deep analytical pipeline.

Quick Start & Requirements

Installation involves cloning the repository, navigating to the oneline directory, and running npm install followed by npm run dev. Configuration requires setting an Exa API Key for news search and selecting an LLM provider (OpenAI or Anthropic) along with corresponding API keys and model choices via the application's settings interface.

Highlighted Details

• Five-Layer Analysis Framework: Provides a structured approach to news intelligence, covering facts, sentiment, context, influence, and risk.
• AI-Powered Insights: Leverages OpenAI or Anthropic APIs for sophisticated natural language understanding and generation.
• Exa Search Integration: Enables smart, API-driven retrieval of relevant news articles for analysis.
• Structured & Exportable Reports: Generates visualized reports in card format, with options to export as Markdown or PDF.
• Source Tracing: Ensures all analytical points are traceable to their original news sources.

Maintenance & Community

The project acknowledges contributions, notably mentioning @snailyp's gemini-balance project for demo API support. Further community links, active development status, or specific maintainer information are not detailed in the provided README.

Licensing & Compatibility

The README does not explicitly state a software license. Users should verify licensing terms before commercial use or integration into closed-source projects. Compatibility with specific operating systems or hardware is not detailed.

Limitations & Caveats

The tool relies heavily on external API keys (Exa, OpenAI/Anthropic), which incur costs and require user provisioning. The effectiveness of the analysis is dependent on the quality of the underlying AI models and the news sources retrieved. Specific limitations regarding unsupported news formats or languages are not mentioned.