Oksskolten

##Functional characteristics - ** Automatic full-text extraction ** -Each article automatically captures the complete content from the source site and extracts it through Readability + 500 noise reduction rules without manual configuration - **AI summary and translation ** -Support Anthropic / Gemini / OpenAI, SSE streaming output, generated based on complete article content - **AI Dialogue ** -Multiple rounds of AI dialogues, supporting MCP tools to search articles and query statistics - ** Full-text search ** -Meilisearch-driven fault-tolerant search covering all article archives - ** Intelligent pulling ** -Adaptive scheduling, ETag/Last-Modified conditional request, content hash deduplication, exponential backoff - **PWA offline reading ** -Supports offline reading, background synchronization, and adding to the main screen - ** Multiple authentications ** -Password + Passkey (WebAuthn) + GitHub OAuth - **14 themes ** -Built-in 14 color themes + JSON import custom themes - ** Article Collection ** -Save any URL as an article and automatically extract the full text - ** Single container architecture ** - API + SPA + Cron scheduler runs in one container Usage method: ###Register an account You need to register an account and password to open the app for the first time ###Add an RSS feed 1. Click "Get Started" in the left sidebar and select **Feeds** 2. Enter website URL or RSS feed address 3. The app automatically discovers RSS feeds and starts pulling articles ###Configure AI functions (optional) See the attached figure 1. Fill in your existing API Key accordingly 2. Ollama can use the address of the warehouse 3. After configuration, select the API Key model required for the function below