vstorm-co/agenticos
Open-source, self-hosted AI agent platform for your company's AI teammates. Build, run and govern agents on your own servers and models: shared chat, RAG, skills, a real browser, 4,700+ MCP servers
About vstorm-co/agenticos
vstorm-co/agenticos is an open-source project on GitHub, mainly written in Python. Open-source, self-hosted AI agent platform for your company's AI teammates. Build, run and govern agents on your own servers and models: shared chat, RAG It currently holds 385 stars and 67 forks with 31 open issues, and was last pushed on 2026-11-17 (repository created 2026-07-31).
Project Overview
Git Homed tracks it on the Today's Trending board, currently at rank #76 with 59 new stars today.
GitHub Repository Details
README
Open-source, self-hosted AI agent platform for your company's AI teammates.
Build, run and govern agents on your own servers and models: shared chat, RAG, skills, a real browser, MCP, budgets, approvals and an audit trail.
Apache-2.0 · built on Pydantic AI
🏆 Built by Vstorm — a recipient of the 2026 Open Agentic AI Innovation Award.
Recognized by the Agentic AI Foundation for adopting AGENTS.md and integrating MCP in its open-source framework.
Read the AAIF announcement · The work behind the award
See it · Highlights · What is it? · Quick start · Server, desktop & mobile · Integrations · Product tour · Find your path · What ships · FAQ · Website · Introduction deck · Documentation
Prepared conversation with fictional sales data in the real Visuals renderer. No live CRM or new model-generation run is represented.
AgenticOS gives every team AI teammates: agents that work with files, run code, browse the web and use your company's tools and knowledge. Your colleagues and your agents work in the same live chat, and any agent can hand code to Claude Code or Codex, on the ChatGPT or Claude subscription you already pay for. It is open source (Apache-2.0) and self-hosted, a layer you run on your own infrastructure. Build and publish agents in the browser, share them with colleagues, and manage their access, cost and results in one place.
New here? Click through the 17-slide introduction: the problem, the idea, the product on real screens, its controls and limits, and how to start. For every screen in detail, open the 47-slide product tour. Arrow keys step through either deck; O lists every slide.
New in the latest releases
- Your local Claude Code or Codex, inside desktop chat. Choose a coding agent per conversation, work in an approved local folder, resume its session and apply corrections while it works. Sign in through the coding agent's own terminal flow.
- Your skills and project instructions. Opt a desktop conversation into local skills and
AGENTS.md/CLAUDE.md, read from your computer. Each new chat starts with this sharing off. - Discord joins Slack, Teams, Telegram and Mattermost. Answer mentions, direct messages and
/ask; review approvals and questions in the channel. Set up a channel. - More useful Visuals and Apps. Draw country maps, pins and routes with the bundled
AOWorldMap, or use the presentation skills to build a deck as an App. A command watcher can follow test or log output while an agent works.
⚡ Highlights
Your team and your agents, in one live room
A project gives everyone the same files, instructions and a computer for the agents. Invite colleagues and work the way a team works: see who is here, watch a teammate's agent answer as it is written, and bring in any agent, tool or document with an @.
Reply to anyone, and the agent knows who said what
Reply to a person or an agent's answer, and the agent applies the change and says whose it was. Every message carries a name, so a decision has an owner.
Claude Code and Codex, on your team
Any agent can hand software work to Claude Code or Codex. They build in the agent's own computer, on your ChatGPT or Claude subscription, and report what they changed, what they checked and what they need from you.
Prepared interface example: Claude Code implements a supplier portal and Codex reviews it in read-only mode. Fictional handoffs; no coding CLI execution is represented.
Ask how it works, get a drawing you can use
An agent draws while it writes: a diagram from your documents, a calculator with sliders, a chart or a 3D scene, live under its message. One click saves it as an app your team opens by link.
A prepared manufacturing Visual inside chat: a supply network, inbound delivery exceptions and assembly windows. Interactive fictional data; no ERP, machine connection or live model call is represented.
Write to an agent and leave
Ask for research that takes a while and close the tab. The agent keeps working, the answer lands in the conversation, and the bell tells you it is done.
Your day, on one screen
Open the console to what your agents did, the calls waiting on you and the conversations to pick up again - with a tour of what they can do that is easy to miss.
Agents that work as a team
A lead agent hands parts of a task to colleagues and writes up what they found - you see who it asked, and what each said.
A real browser, right in the chat
Watch it click through sites, take over when a sign-in needs you, point at a spot to comment.
Build, equip and test in one window
Enable tools and try a real task in the draft test panel before publishing the agent for colleagues.
Apps, not just answers
An agent publishes a dashboard or a report as a page with a stable link and versions.
Northline: a prepared supply and production planning App with fictional shipments, local delay simulation and assembly windows.
More prepared Apps: motor assembly and sales intelligence
Working local demonstrations with fictional data. Motor Atlas lets you inspect a conceptual assembly and its component records; Revenue Pulse recalculates sales metrics by region and period. No equipment, ERP or CRM is connected.
A website widget, on your own site
The actual AgenticOS widget runs a published, read-only agent on the fictional Trail Supply website. In the recorded local test it answered using the seeded catalog in its instructions. Live inventory, purchases and payments are not connected.
Build and publish a website widget
4,700+ MCP servers
Connect Notion, GitHub, Jira or your own server, and choose which tools each agent may call.
Work in AgenticOS from Claude Code, Codex or Cursor
AgenticOS is an MCP server. Add it with one command and sign in through the browser: your coding assistant can list and run agents, read runs and what they cost, and search and fill knowledge bases, within what you allowed.
📸 See it in action
Browse the current product gallery. Motor Atlas below is a prepared, interactive assembly explorer with fictional component records and telemetry; it is not a CAD drawing.
🧩 Build useful work around your agents
Use Workflows to connect typed steps, agent calls and MCP tools on a visual canvas. Store structured results in Virtual Tables, start work with Routines, and review runs and approvals. Voice calls and the desktop app provide other ways to work with the same published agents.
Visual workflows, from lead intake to a human-reviewed handoff.
Illustrative 30-step draft with Google Workspace, Microsoft 365, HubSpot, Salesforce, Notion, Slack, agents and approvals. Connections are demo placeholders; this is not an executed integration.
Structured work your agents and your team can read together.
Virtual Tables with fictional companies, typed fields, stages and next actions. Switch between a grid and a kanban board.
Start with a recognizable task: an LLM wiki in a Project, a website agent widget, or a team working from shared project files and instructions. A widget supplies the conversation surface; recommendations and cart actions still need your catalog, tool integration and permission checks.
💡 What is AgenticOS?
AgenticOS is an open-source (Apache-2.0), self-hosted layer for building, sharing and governing AI agents across a company. Teams configure an agent in the browser by writing its instructions, choosing a model and switching on tools. They connect it to company documents and apps, and publish it to web chat, Slack, Teams, Discord, Mattermost, Telegram, a website widget or an API. Administrators control who can use each agent, what it may spend and which actions need a person's approval. Every run is recorded.
Most agent frameworks give you a library, so every change to an agent's behaviour is a pull request, a review and a release. That is the wrong shape for the small agents a company actually wants, because the person who knows what the agent should say is not the person with commit access. Code defines, configuration composes: engineers extend what there is to assemble, and configuration can only ever reach what code registered.
It runs on your own infrastructure with Docker Compose and works with 28 model providers, including your ChatGPT subscription and local models through Ollama and vLLM. Agents run on Pydantic AI and pydantic-ai-harness; the layer around them uses FastAPI, PostgreSQL with pgvector and Next.js. It is maintained by Vstorm.
Who it is for:
- Companies that want an internal AI agent layer they own, instead of per-seat assistants in a vendor's cloud.
- Teams with repeated work over documents and tools, such as reports, support answers, contract checks and data analysis.
- IT and security teams that need data sovereignty, company sign-in, budgets, approvals and an audit trail for AI agents.
- Engineers who want typed Python extension points and a console their non-technical colleagues can use.
How it fits into your company. Illustrated people; the agents are examples.
How AgenticOS fits into a company: departments such as finance, operations or the board use shared agents in web chat, Slack or a private workspace. Your systems call agents through the API, and events or schedules start them automatically. Every request passes the same controls: roles, budgets, approvals, guardrails and a recorded run. Your data, vectors, code sandboxes, the credential vault and optional local models stay on your infrastructure. Hosted models, SaaS tools and document sources outside are used only when you configure them.
28 model providers · 4,700+ MCP gallery entries · shared chat and Projects · visual workflows and virtual tables · web, desktop, API and messaging channels. Configure availability and tools for each published agent.
✨ What you can do
| For your team | What AgenticOS provides | |---|---| | Work with files and code | Analyze CSVs, produce charts and documents, work on repositories | | Build reusable agents | Choose models and tools, publish versions, share agents with colleagues | | Connect company knowledge | Reuse skills, context and searchable documents across agents | | Share the results | Publish interactive pages with stable links and version history | | Run and monitor | Customize dashboards, inspect runs, schedule tasks and set budgets | | Organize company access | Combine roles, department groups and company sign-in |
🚀 Quick start
Choose how you want to start
One company workspace, many ways to connect. The server runs AgenticOS; the desktop app connects to it. Joining an existing workspace needs no Docker or developer tools.
| Your situation | Next step | |---|---| | Your company already has AgenticOS | Ask for the workspace URL and sign in through your browser, or install the desktop app. | | You want a workspace for your company | Have your IT team deploy a server, or ask Vstorm to help. | | You want to try it on your computer without terminal commands | Install the desktop app, choose On this computer and follow setup. Review the requirements first; AI requires a configured model, while recorded examples need none. | | You are a developer exploring locally | Run the Docker quickstart. |
Try it locally
For developers: try AgenticOS locally with Docker
Start with Docker Compose and access to a model provider. On macOS or Linux, run:
curl -fsSL https://raw.githubusercontent.com/vstorm-co/agenticos/main/scripts/quickstart.sh | bash
On Windows, run the same command inside WSL2 with Docker Desktop's WSL2 integration switched on. The installer asks for a model provider and key, your login and an organization name, pulls the published images and starts a deployment with a working agent in it. A host with 4 vCPU and 8 GB of RAM runs it. Then open http://localhost:3000 and sign in with the login you chose during installation.
Your first agent: follow the document-assistant walkthrough to upload a handbook, ask questions and check answers against cited sources. Then pick a next task from 30 tutorials, each with a sample input and a check you can run.
Inspect the installer or deploy another way
Read the installer before running it. To check prerequisites without installing:
curl -fsSL https://raw.githubusercontent.com/vstorm-co/agenticos/main/scripts/quickstart.sh | bash -s -- --check
For manual Docker Compose setup, pinned versions and troubleshooting, follow the installation guide. For development from source, see Contributing.
☁️ One-click deploy
Review the provider's instructions, image versions and included services before deploying. These hosting templates are maintained separately from this repository. The Railway template currently pins an older AgenticOS release and does not include the local code sandbox.
Run on your server. Open it anywhere.
The server holds the workspace. The browser, desktop app and phone connect to that deployment with its accounts, agents and access controls.
Deploy a company server
Your IT team sets up the workspace once, then shares an HTTPS address and creates your account. You can use that address in a browser or in the desktop app. For an assisted rollout, talk to Vstorm.
Server setup for IT and developers
Use a Linux host with 4 vCPU and 8 GB RAM, Docker Engine 24+, Compose 2.24+ and make. Point two DNS names at it — for example app.example.com and api.example.com — and allow HTTPS through a reverse proxy. The supported deployment is Docker Compose on one host.
# On the server. Start from a released version so code and images stay aligned.
git clone --branch v1.38.2 --depth 1 https://github.com/vstorm-co/agenticos.git
cd agenticos
bash scripts/server-init.sh
The setup asks for the site/API hostnames and certificate email, then generates backend/.env. Set AGENTICOS_VERSION=1.38.2 there to pin this release. For a new host using the bundled Traefik proxy:
docker network create traefik_webgateway
docker compose --env-file backend/.env -f docker-compose-traefik.yml up -d
make prod PROXY=traefik
make prod-frontend PROXY=traefik
Then bootstrap the first organization, owner and agent, open your site's HTTPS address and add your organization's model credentials. The guide also covers an existing proxy, Nginx, updates and backups. Keep a backup of the database and the vault key in backend/.env.
```bash docker compose --env-file backend/.env -f docker-compose-prod.ym

















