File system memory for artificial intelligence agents
Developed by Volcengine, OpenViking offers a self-improving context database for AI agents. This system combines agent memory, information retrieval (RAG) processes and abilities under a single roof.
Updates
- October 23, 2026: Stars 40,425 → 40,663, latest release v0.5.2 (October 23, 2026).
- October 20, 2026: Stars 40,333 → 40,425, latest release v0.5.1.1 (October 20, 2026).
- October 19, 2026: Stars 39,519 → 40,333, latest release v0.5.1 (October 19, 2026).
- October 9, 2026: Stars 39,151 → 39,519, latest release v0.5.0 (October 9, 2026).
What you get
- Organizes information hierarchically like a file system.
- It reduces the cost of artificial intelligence with layered loading.
- Makes agent history traceable and debuggable.
Installation
pip install openviking --upgrade
openviking-server init # interactive wizard: providers, models, ov.conf
openviking-server doctor # validate setup
openviking-server # start (background: nohup openviking-server > openviking.log 2>&1 &)Running it
pip install "openviking[bot]"
openviking-server --with-bot
ov chat # in another terminalIf you don't write code
Construct context management for an artificial intelligence agent using the OpenViking database. It structures the information via the viking:// protocol by separating the information into L0 summary, L1 overview and L2 detail layers. By placing the agent's memory, resources, and capabilities in this virtual file system, it allows it to navigate directories during interrogation and create long-term memory by learning from past sessions.
Repository: https://github.com/volcengine/OpenViking. Follow the official instructions in the README of this repository for setup and execution.
Before starting, check which OS we are on and if the necessary tools (e.g., Python, Node.js, Git, Docker) are installed. Install any missing ones step-by-step according to this system.
Create a new, empty folder to work in. Do not touch any files outside this folder and do not delete anything. Tell me what you are going to do in one sentence before running each command.
Tell me what I should see on the screen at the end of each step. If an error occurs, I will paste the error message to you exactly as it is, and we will fix it together.
Related dictionary terms
Links
TreScout did not build this tool · we found it in GitHub trends and wrote it up. This page describes the repository as of 2026-08-18: The star count and our text belong to that day, the repository may have changed since. Check the repository link for the current state. The introductory text was prepared with AI from the repository's own documentation, and this page was machine-translated from the Turkish original · the Turkish version prevails.