MineContext is an open-source, proactive AI assistant designed to capture, understand, and leverage a user’s digital context in order to provide meaningful insights, summaries, and productivity support. The system continuously collects contextual data from sources such as screenshots and user activity, then processes and organizes this information into structured knowledge that can be reused later. Unlike traditional chat-based assistants, MineContext operates in the background and delivers proactive outputs such as daily summaries, task suggestions, and contextual reminders without requiring explicit prompts. It is built around a context engineering framework that manages the full lifecycle of data, including capture, processing, storage, retrieval, and consumption. The platform emphasizes privacy through a local-first architecture, allowing users to keep their data stored and processed on their own device rather than relying on external cloud services.
Features
- Proactive AI assistant that generates summaries todos and insights automatically
- Context capture from screenshots documents and user activity
- Local first architecture ensuring privacy and on device data processing
- Support for multiple AI models including local and API based models
- Context engineering pipeline covering capture processing storage and retrieval
- Intelligent resurfacing of relevant information during workflows