OpenVole
The self-hosted agent OS — run a fleet of AI agents on your own hardware, against any model, from one dashboard. Then link them across machines into a distributed VoleNet mesh — nodes that share tools, memory, and brains.
vole serve
One vole serve control plane managing a fleet of agents — each its own isolated space.
🏠 Self-hosted & private
Your agents, your hardware, your data. Point a space at local Ollama and nothing touches the cloud.
🔌 Model-agnostic
One brain paw for Anthropic, OpenAI, Gemini, xAI, and Ollama. Switch per agent with one env var.
🖥️ A server, not a script
vole serve is a control plane — create, run, and chat with a whole fleet of agents from one dashboard.
🕸️ VoleNet
A peer-to-peer agent mesh — remote tools become local, workers share one brain, memory syncs. No central server.
🔐 Sandboxed & signed
Every paw runs in a locked-down subprocess with explicit permissions. Every VoleNet action is cryptographically signed and authorized before it runs — nothing executes on trust alone.
🛡️ Post-quantum ready
Mesh identities are signed with hybrid Ed25519 + ML-DSA-65 (FIPS 204), so your agent network stays verifiable even against tomorrow's quantum attacks.
🧩 Microkernel, zero lock-in
A tiny, LLM-ignorant core. Every capability is a Paw you can swap, sandbox, or write yourself.
🪟 An app platform
A Paw can ship its own UI, rendered under the dashboard's Apps tab — extend it into a full app.