OpenViking
AIOpenViking is an open-source context database that stores agent memories, resources, and skills as one browsable virtual filesystem.
OpenViking
OpenViking is an open-source context database that stores agent memories, resources, and skills as one browsable virtual filesystem.
About OpenViking
OpenViking is an AGPLv3 open-source context database for AI agents that replaces the opaque vector store with a virtual filesystem. Memories, resources, and skills each get a viking:// URI, so an agent locates and manipulates its own context deterministically using ls, tree, and find the way a developer works with files. Every entry is processed on write into three tiers — an L0 one-sentence abstract, an L1 overview of core information and usage scenarios, and L2 full original detail — and loaded only as deep as the task requires, which cuts token spend substantially. Retrieval works by directory recursion: vector search first finds the highest-scoring directory, then drills down layer by layer so results arrive with their surrounding context intact, and each query preserves its browsing trajectory so a wrong result can be traced to the exact path that produced it. After a session commits, OpenViking asynchronously extracts user preferences and agent experience into long-term memory. Published benchmarks on LoCoMo and tau2-bench report large accuracy gains — for example Claude Code moving from 57.21% to 80.32% on LoCoMo.
Key Features
Use Cases
Frequently asked questions about OpenViking
What is OpenViking?
OpenViking is an open-source context database that stores agent memories, resources, and skills as one browsable virtual filesystem.
How does OpenViking work?
OpenViking works by combining Viking:// Virtual Filesystem: Memories, resources, and skills each receive a URI in one unified namespace, so agents browse context with ls, tree, and find instead of querying a black-box store., Three-Tier Context Layers: Each entry is written as an L0 abstract, L1 overview, and L2 full detail, letting an agent judge relevance cheaply and load full data only when needed., Directory Recursive Retrieval: Vector search locates the highest-scoring directory first and then descends layer by layer, so retrieved fragments keep their surrounding context., Observable Retrieval Trajectories: Every query records the directory-browsing path it took, so an incorrect result can be traced back to the exact decision that produced it., Sessions Become Memory: After a session commits, user preferences and agent experience are asynchronously extracted into long-term memory without blocking the agent. to help users with Long-Term Agent Memory: Give a coding or assistant agent persistent recall of user preferences and past sessions across long-running conversations., Reducing Token Spend: Teams paying for oversized context windows load L0 abstracts for triage and pull L2 detail only for the entries that matter., Debugging Bad Retrievals: Engineers inspect the recorded browsing trajectory to find out why an agent surfaced the wrong document instead of guessing at embedding behavior., Knowledge Base Question Answering: Serve structured organizational knowledge to agents with directory-level context preserved around every answer., Skill and Resource Management: Store reusable agent skills alongside memories and documents in one addressable namespace instead of separate systems..
What are the main features of OpenViking?
Key features include Viking:// Virtual Filesystem: Memories, resources, and skills each receive a URI in one unified namespace, so agents browse context with ls, tree, and find instead of querying a black-box store., Three-Tier Context Layers: Each entry is written as an L0 abstract, L1 overview, and L2 full detail, letting an agent judge relevance cheaply and load full data only when needed., Directory Recursive Retrieval: Vector search locates the highest-scoring directory first and then descends layer by layer, so retrieved fragments keep their surrounding context., Observable Retrieval Trajectories: Every query records the directory-browsing path it took, so an incorrect result can be traced back to the exact decision that produced it., Sessions Become Memory: After a session commits, user preferences and agent experience are asynchronously extracted into long-term memory without blocking the agent..
Who is OpenViking for?
OpenViking is useful for anyone interested in Long-Term Agent Memory: Give a coding or assistant agent persistent recall of user preferences and past sessions across long-running conversations., Reducing Token Spend: Teams paying for oversized context windows load L0 abstracts for triage and pull L2 detail only for the entries that matter., Debugging Bad Retrievals: Engineers inspect the recorded browsing trajectory to find out why an agent surfaced the wrong document instead of guessing at embedding behavior., Knowledge Base Question Answering: Serve structured organizational knowledge to agents with directory-level context preserved around every answer., Skill and Resource Management: Store reusable agent skills alongside memories and documents in one addressable namespace instead of separate systems..
How much does OpenViking cost?
OpenViking is free to use.
How do I get started with OpenViking?
Visit https://github.com/volcengine/OpenViking to sign up and explore OpenViking.
Explore more AI Ai Agents tools
Browse all Ai Agents tools →Compare OpenViking: vs Apache Maka · vs Construct Computer · vs ShogunAI · vs fx