OpenSpry is an agentic operating system that runs 24/7 on your machine. It doesn't just answer questions — it manages a living graph of your intentions, dispatching AI agents to advance them while you sleep.
OpenSpry is an operating system for your goals. A persistent kernel that works continuously, not just when you're talking to it.
A clock ticks every 60 seconds, recalculating priorities and dispatching agents — whether or not you're at the keyboard.
Your goals form a living DAG with decomposition, dependencies, and tensions. The system tracks how everything connects.
Type what you need in plain English. OpenSpry classifies it, creates intents, runs agents, and reports back.
Built-in web search and shell. Install community tools for email, calendars, messaging — or build your own in 15 minutes.
Four-tier hierarchy from working context to permanent record. Everything agents learn is stored, searchable, and builds over time.
Your data stays on your machine. No cloud account, no telemetry. OpenSpry runs as a local server with a browser dashboard.
OpenSpry operates like an OS for your intentions. Here's the loop.
In natural language via the web dashboard or CLI. "Finish the report by Friday" or "Keep my inbox clean."
Your words become nodes in a living graph — with priorities, deadlines, dependencies, and constraints.
Every 60 seconds, the kernel recalculates priorities and emits signals for whatever needs attention most.
LLM-powered agents pick up signals, use tools to research and execute, then write evidence back to the graph.
Results, decisions, and observations are stored in persistent memory — building richer context over time.
Not locked in. Switch providers at any time from the chat — no restart required.
GPT-4o, GPT-4o-mini
set key openai sk-...
Claude Sonnet, Opus
set key anthropic sk-ant-...
Gemini 2.0 Flash
set key google AIza...
Local, free, private
set key ollama local
Whether you're managing a solo business, orchestrating a side project, or just want your computer to actually do things for you.
Automate the operational overhead so you can focus on building.
Build tools, extend the ecosystem, integrate with your stack.
An AI that works while you sleep, remembers everything, and gets smarter.