context-use

context-use is a command-line tool designed to transform exported data from various search, social, and AI applications into searchable memory stores for use with AI agents. Users can download their data from supported providers such as ChatGPT, Instagram, Claude, Google, and Facebook, then process it locally to create rich, persistent memories that can be queried or used as context by personal AI agents. The tool is intended for individuals who want to leverage their personal data as context for AI-driven workflows.

The platform offers both quick and full data processing pipelines. The quick pipeline uses a real-time API to extract memories from the most recent 30 days of exported data, which is suitable for smaller datasets but may encounter rate limits with larger exports. The full pipeline utilizes a batch API, making it more cost-effective and less susceptible to rate limits, with typical processing times ranging from 2 to 10 minutes. Extracted memories are stored in a local SQLite database, ensuring persistence across sessions.

Users can interact with their memory stores through commands to list, search, and export memories. Additionally, context-use provides a multi-turn personal agent capable of synthesizing information, profiling, and answering questions based on the user's full memory store. Configuration is managed via a configuration file saved in the user’s home directory. The tool is open source, portable, and runs locally on the user's machine.

Future features include expanded support for additional data sources such as LinkedIn, TikTok, Pinterest, Amazon, and Booking. There is also an upcoming managed cloud version, context-use cloud, which promises web consent flows, regular data synchronization, and dedicated support. , with operations in Europe and the US.