01Automated research pipeline (search, ingest, bundle, upload, generate, download) with a single command or AI agent call.
02Compatibility with a wide range of AI hosts (Claude, Cursor, ChatGPT, Gemini, etc.) via MCP server, REST API, or subprocess calls.
03Interactive HTTP dashboard for overview, library management, diagnostics, and CLI actions as buttons.
04Integration with Obsidian, generating structured Markdown notes for each paper and cluster-specific Bases dashboards.
05Generation of pre-computed 'crystals' (canonical answers) and a structured memory layer (entities, claims, methods) for efficient knowledge retrieval.
064 GitHub stars