Acontext is an LLM orchestration backend and agent memory framework designed to manage session state and knowledge for AI agents. It functions as a context manager and orchestration layer that integrates model providers with a secure code sandbox and a zero-knowledge data store.
The project is distinguished by its approach to knowledge distillation, capturing agent learnings as reusable Markdown skills and structured memory files. It provides a secure execution environment where shell commands and scripts run in isolated containers with the ability to mount these persistent skill files directly into the sandbox.
The system covers a broad range of capabilities, including context window optimization through summarization and compression, multi-tenant resource isolation, and encrypted data persistence. It also includes tools for autonomous agent control, such as iteration limiting and automatic task extraction from conversations.
The runtime can be executed locally via a command-line interface and includes support for hosting a local memory server and database.