explaingit

kortix-ai/suna

📈 Trending19,764TypeScriptAudience · pm founderComplexity · 4/5ActiveSetup · moderate

TLDR

An open-source OS for running multiple AI agents on a shared Linux machine that automate business operations by accessing files, APIs, and tools together.

Mindmap

mindmap
  root((Suna))
    What it does
      Shared Linux environment
      Multiple AI agents
      Automate operations
    How it works
      Agents share filesystem
      Shared credentials
      Work history access
    Use cases
      Support automation
      Finance processing
      Sales workflows
      Operations tasks
    Tech stack
      TypeScript
      Linux
      OpenCode runtime
    Getting started
      Single curl install
      CLI management
      Local or server

Things people build with this

USE CASE 1

Automate customer support by having an AI agent handle tickets and share context with other agents.

USE CASE 2

Run finance workflows where an invoice-processing agent updates shared records that sales and operations agents can access.

USE CASE 3

Operate a 24/7 autonomous team handling routine business tasks like email, API calls, and document processing without human intervention.

USE CASE 4

Build specialized agents for sales, support, and operations that collaborate through a shared environment and memory.

Tech stack

TypeScriptLinuxOpenCodeBashNode.js

Getting it running

Difficulty · moderate Time to first run · 30min

Requires Linux environment and Node.js runtime; multi-agent coordination setup needs configuration.

License could not be detected automatically. Check the repository's LICENSE file before use.

In plain English

Suna (now branded as Kortix) is an open-source operating system for running autonomous AI agents that can handle the ongoing work of operating a company. The central concept is a shared Linux machine, running 24 hours a day, 7 days a week, where multiple AI agents all share the same filesystem, databases, credentials, and accumulated work history. Because all agents share the same context, information flows naturally between them: when a support agent resolves a customer ticket, the product agent already has that context; when a finance agent processes an invoice, every other agent knows about it. The underlying idea is that coding agents, AI systems that operate inside a real Linux environment with access to bash, file systems, APIs, and the full command-line toolchain, are general-purpose workers, not just software engineering assistants. The same capability that lets an agent write code also lets it call APIs, parse documents, generate reports, manage databases, send emails, and browse the web. By giving specialized agents (for sales, support, finance, operations) access to the same shared environment and memory, the system aims to automate a large fraction of business operations without human involvement for routine tasks. You install it with a single curl command on either a local machine or a server, and manage it through a CLI with commands to start, stop, update, and check status. The runtime layer is OpenCode. The project is written in TypeScript and is open source.

Copy-paste prompts

Prompt 1
How do I install Suna on my server and set up multiple AI agents to share the same Linux environment?
Prompt 2
Show me how to create a support agent and a finance agent that both access the same shared filesystem and credentials in Suna.
Prompt 3
What CLI commands do I use to start, stop, and monitor AI agents running on Suna?
Prompt 4
How can I configure a Suna agent to call APIs, parse documents, and send emails as part of an automated workflow?
Prompt 5
Explain how agents in Suna share work history and context so that information flows between support, sales, and operations teams.
Open on GitHub → Explain another repo

Generated 2026-05-18 · Model: sonnet-4-6 · Verify against the repo before relying on details.