Origin is a Rust-based daemon that provides a local-first approach to managing AI agent memory and context. It integrates Git for versioning memories, allowing developers to track, inspect, and revert changes to an agent's knowledge base. The system also distills raw memories into wiki pages, making the agent's learned context human-readable and auditable. Origin supports various AI clients, including Claude Code, Cursor, Codex, and any Model Context Protocol (MCP) client, facilitating consistent context across different AI development environments. Key features include composition of memories into pages, session tracking, an entity graph for linking related information, and hybrid retrieval combining vector, full-text search, and graph-based recall. The daemon enforces provenance by requiring wiki pages to cite source memory IDs and supports background cycles to refresh pages as new memories arrive. It also flags low-confidence captures and contradictions for review, enhancing the trustworthiness of the agent's memory. The latest release, v0.7.0, introduces cross-platform Linux and Windows support, along with new evaluation benchmarks for KG-faithfulness and page-distillation faithfulness.
Origin: A Local-First Rust Daemon for AI Agent Memory and Context Management
Origin is a local-first Rust daemon designed to manage AI agent memory and context. It features Git-versioned memories, distilled wiki pages, and supports sessions for various AI clients like Claude Code, Cursor, and Codex, aiming to provide persistent context across AI workflows.
Why it matters
Origin addresses a key challenge in AI agent development: maintaining consistent context and memory across sessions. By offering Git-versioned memory and a system for distilling information into wiki pages, it provides a robust, auditable, and local-first solution for long-term AI agent memory, potentially enhancing the reliability and continuity of AI-driven development.