Meet OpenViking: An Open-Source Context Database that Brings Filesystem-Based Memory and Retrieval to AI Agent Systems like OpenClaw

OpenViking: An Open-Source Context Database for AI Agents

OpenViking is an open-source Context Database for AI Agents from Volcengine. The project is built around a simple architectural concept: agent systems should not treat context as a flat collection of text chunks. Instead, OpenViking organizes context through a file system paradigm, with the goal of making memory, resources, and skills manageable through a unified hierarchical structure.

A Virtual Filesystem for Context Management

At the center of the design is a virtual filesystem exposed under the viking:// protocol. OpenViking maps different context types into directories, including resources, user, and agent. Under those top-level directories, an agent can access project documents, user preferences, task memories, skills, and instructions.

Directory Recursive Retrieval

The project's retrieval pipeline first uses vector retrieval to identify a high-score directory, then performs a second retrieval within that directory, and recursively drills down into subdirectories if needed. The README calls this Directory Recursive Retrieval. The basic idea is that retrieval should preserve both local relevance and global context structure.

Tiered Context Loading

OpenViking also adds a built-in mechanism for Tiered Context Loading. When context is written, the system automatically processes it into three layers. L0 is an abstract, described as a one-sentence summary used for quick retrieval and identification. L1 is an overview that contains core information and usage scenarios for planning. L2 is the full original content, intended for deep reading only when necessary.

Retrieval Observability and Debugging

A second important systems feature is observability. OpenViking stores the trajectory of directory browsing and file positioning during retrieval. The README file describes this as Visualized Retrieval Trajectory. In practical terms, that means developers can inspect how the system navigated the hierarchy to fetch context.

Session Memory and Self-Iteration

The project also extends memory management beyond conversation logging. OpenViking includes Automatic Session Memory and Self-Iteration, which allows agents to learn from their own experiences and adapt to new situations.