HyperAgency is an Agentic AI Operating System that replaces brittle, fragile automations with long-lived, self-improving systems. HyperAgency allows organizations to deploy autonomous agents that remember, adapt, and act on your behalf, under human governance and auditability.
It is open-source, model-agnostic, and cloud/self-hosted, providing the infrastructure to build fully autonomous AI-human organizations.
🧠 Think of HyperAgency as your operating system for autonomous workflows — distributed, flexible, and composable.
📘 Explore the HyperAgency Documentation for full capabilities, architecture, agent archetypes, integration examples, and step-by-step guides.
📄 Read the HyperAgency Research Paper for the system design, implementation and use cases.
HyperAgency is more than an AI platform — it's a new operational layer:
- Persistent Agency — agents maintain memory and context across time, adapting to real-world changes.
- Coordinated Intelligence — multiple agents work together to solve complex problems.
- Human Governance — clear approval points, credential management, and auditable behavior.
- Omni-Channel Connectivity — integrate across messengers, web apps, email, and APIs.
- Decentralized Architecture — connect multiple HyperAgency instances for distributed collaboration.
- Web3 Marketplace — publish, buy, and monetize agentic workflows and HyperAgents.
- MetaAgent — self-developing AI that creates, evolves, and orchestrates agents autonomously.
HyperAgency turns AI from fragile scripts and prompts into living, compound intelligence systems.
- ✅ Create & Deploy Agents — instant archetypes or custom blueprints.
- 💬 Multi-Agent Communication — human and AI interaction in one interface.
- 🗺️ Map View — visualize workflows and agent coordination.
- 🧩 Composable Systems — agents can orchestrate other agents, integrate APIs, run scripts, generate content, or process data.
- 🔀 Persistent Memory — knowledge, workflow state, and context are shared and versioned.
- 🔐 Governance & Security — Vault-secured credentials, audit logs, approvals.
- 🌍 Distributed Nodes — federated network for scalable collaboration.
- ⚡ Rapid Deployment — cloud or local containerized setup, production-ready.
HyperAgency comes with 20+ production-ready agents you can deploy instantly:
| Archetype | Role / Function |
|---|---|
| Chat | Natural language communication |
| RAG | Research & document retrieval |
| ImageGen | Image generation from prompts |
| BrowserUse | Web navigation & task automation |
| Hermes | Autonomous general-purpose agent |
| OpenClaw | Personal AI assistant |
| Codex | Software engineering agent |
| ClaudeCode (connector) | Coding assistant connector |
| STT | Speech-to-text transcription |
| TTS | Text-to-speech synthesis |
| Avatar | Talking faces with lip-sync |
| Maptrix | Nested workflow execution |
| Code | Python, JavaScript, Bash computation |
| Command | Linux command execution |
| Quantum | Quantum computing tasks |
| Storage | Key-value and file storage |
| Transform | Text transformation & templating |
| System | HyperAgency infrastructure management |
| MCP | Tool execution via Model Context Protocol |
| A2A | Agent-to-agent delegation |
| Curl | API / HTTP requests |
| Langflow | AI strategy & multi-agent orchestration |
| Node-RED | External API & IoT workflow integration |
| N8n | Flexible automation workflows |
| Notebook | Parameterized Jupyter notebooks |
All agents are composable, versionable, portable (JSON), auditable, and reproducible.
Each agent can run any compatible LLM or model based on role, cost, latency, or compliance:
- Ollama (local open-source LLMs)
- Speaches (local open-source TTS/STT models)
- OpenAI
- Anthropic
- Google (Vertex AI / GenAI)
- Azure
- Mistral
- Groq
- Cohere
- Fireworks
- Together
- Hugging Face
- NVIDIA
- Amazon / Bedrock
- Databricks
- IBM
- AI21
- Upstage
- xAI
- Perplexity
Agents are bound to intent, not to a single model — no vendor lock-in.
Deploy agents across any communication channel in seconds:
- Messengers: XMPP, Telegram, Slack, Discord, MS Teams, WhatsApp, VK, Mattermost, Matrix, IRC, Zulip, Rocket Chat, SSH Chat, NextCloud Talk, Muble, Keybase
- Email: IMAP/SMTP
- Phone: FreeSwitch
- Scheduler: Cron
- Webhooks Server
- A2A Server
- MCP Server
- Web App: Lowdefy
- XMPP Server
Bridges let agents communicate, automate, and act across your digital ecosystem.
Full visibility into agent actions, memory, and evolution:
- Logs & metrics in real-time
- Debug and trace workflows across agents
- Performance monitoring and system health
- Privacy-first design — isolated data per organization
HyperAgency supports secure peer-to-peer networks:
- Scale horizontally with multiple nodes
- Share agents across trusted nodes and teams
- Retain full control over privacy and data ownership
- Publish, share, and monetize HyperAgent packages
- Join a global ecosystem of builders and innovators
- Enable agentic deals with verifiable capabilities
This section helps you understand what capabilities are available depending on how you install HyperAgency.
We support four deployment modes:
- A. 🧩 Standalone (individual) — single all-in-one Docker container for quick start and local testing
- B. 🐳 Docker Cluster (team) — multi-service Docker Compose setup for personal or team deployments
- C/D. ☸️ Kubernetes Cluster (enterprise) — scalable enterprise-grade infrastructure deployment (on-prem or private-cloud)
- E. ☁️ HyperAgency Cloud (any) — fully managed hosted platform at https://h9y.ai
| Feature | Standalone | Docker Cluster | Kubernetes Cluster | HyperAgency Cloud |
|---|---|---|---|---|
| Web-Landing | ❌ | ❌ | ✅ | ✅ |
| Dashboard | ✅ | ✅ | ✅ | ✅ |
| Hive | ✅ | ✅ | ✅ | ✅ |
| Chat | ✅ | ✅ | ✅ | ✅ |
| Map | ✅ | ✅ | ✅ | ✅ |
| Data | ✅ | ✅ | ✅ | ✅ |
| Observability | ❌ | ✅ | ✅ | ✅ |
| Omni-channel | ✅ | ✅ | ✅ | ✅ |
| Apps | ✅ | ✅ | ✅ | ✅ |
| Wallet | ❌ | ❌ | ✅ | ✅ |
| Vault | ❌ | ✅ | ✅ | ✅ |
| CLI | ✅ | ✅ | ✅ | ✅ |
| API & SDK | ✅ | ✅ | ✅ | ✅ |
| Mobile App | ❌ | ❌ | ✅ | ✅ |
| Agent | Standalone | Docker Cluster | Kubernetes Cluster | HyperAgency Cloud |
|---|---|---|---|---|
| Chat | ❌ | ✅ | ✅ | ✅ |
| RAG | ❌ | ✅ | ✅ | ✅ |
| ImageGen | ✅ | ✅ | ✅ | ✅ |
| BrowserUse | ❌ | ✅ | ✅ | ✅ |
| OpenClaw | ❌ | ✅ | ✅ | ✅ |
| Hermes | ❌ | ✅ | ✅ | ✅ |
| Codex | ❌ | ✅ | ✅ | ✅ |
| Claude Code | ❌ | ❌ | ❌ | ✅ |
| STT | ❌ | ✅ | ✅ | ✅ |
| TTS | ❌ | ✅ | ✅ | ✅ |
| Avatar | ❌ | ✅ | ✅ | ✅ |
| Maptrix | ✅ | ✅ | ✅ | ✅ |
| Code | ❌ | ✅ | ✅ | ✅ |
| Command | ✅ | ✅ | ✅ | ✅ |
| Quantum | ❌ | ✅ | ✅ | ✅ |
| Storage | ✅ | ✅ | ✅ | ✅ |
| Transform | ✅ | ✅ | ✅ | ✅ |
| System | ✅ | ✅ | ✅ | ✅ |
| MCP | ✅ | ✅ | ✅ | ✅ |
| A2A | ✅ | ✅ | ✅ | ✅ |
| Curl | ✅ | ✅ | ✅ | ✅ |
| Langflow | ❌ | ✅ | ✅ | ✅ |
| Node-RED | ❌ | ✅ | ✅ | ✅ |
| N8n | ❌ | ✅ | ✅ | ✅ |
| Notebook | ❌ | ✅ | ✅ | ✅ |
| Bridge | Standalone | Docker Cluster | Kubernetes Cluster | HyperAgency Cloud |
|---|---|---|---|---|
| Messengers | ❌ | ✅ | ✅ | ✅ |
| ✅ | ✅ | ✅ | ✅ | |
| Phone | ❌ | ❌ | ✅ | ✅ |
| Scheduler | ✅ | ✅ | ✅ | ✅ |
| Webhooks Server | ✅ | ✅ | ✅ | ✅ |
| A2A Server | ✅ | ✅ | ✅ | ✅ |
| MCP Server | ✅ | ✅ | ✅ | ✅ |
| Web App (Lowdefy) | ❌ | ✅ | ✅ | ✅ |
| XMPP Server | ✅ | ✅ | ✅ | ✅ |
If you don’t want to manage infrastructure, just use the hosted version of HyperAgency Cloud (SaaS Self-Serve).
👉 HyperAgency Cloud (SaaS) https://h9y.ai
No Docker, no Kubernetes, no setup — just create an account and start using agents immediately.
You can self-host HyperAgency locally in just a few steps — or skip the setup entirely and use our cloud environment for a faster start.
To run HyperAgency locally, install the following:
Quick start:
docker run \
-p 3990:3990 -p 6369:6369 -p 6370:6370 -p 5280:5280 \
-v h9y-standalone-volume:/var/lib/ \
ghcr.io/vuics/h9y-standalone:v1.39.0
# Or use latest image:
# ghcr.io/vuics/h9y-standalone:latestOpen localhost:3990 in the browser and start using standalone h9y.
mkdir hyperagency
cd hyperagency/
curl -fsSL https://raw.githubusercontent.com/vuics/h9y/refs/heads/main/install.sh | sh
# RECOMMENDED: Generate TLS certificates.
# (Otherwise, use your own certificates.)
./gen-certs.sh
# OPTIONALLY: Setup DNS with /etc/hosts
./setup-hosts.shRun:
docker compose upAfter you started the stack, you can open in the browser:
| App Name | URL |
|---|---|
| App | h9y.localhost |
| API | api.h9y.localhost |
| Vault | vault.h9y.localhost |
| Langflow | langflow.h9y.localhost |
| Nodered | nodered.h9y.localhost |
NOTE: Replace h9y.localhost with your ${DOMAIN}.
- Open Vault (replace
h9y.localhostwith your${DOMAIN}), and input: • Key shares:1• Key threshold:1 - Set the env vars with displayed initial root token and the keys in the
.envfile in the format:
VAULT_TOKEN=(Initial root token)
VAULT_UNSEAL_KEYS=(Key 1)- Unseal the vault by inputing the key, and sing into vault with the initial root token.
- Enable new engine with type KV (kv-v2) and path
secret. - Restart Docker Compose.
docker compose down
docker compose upHyperAgency ships with predefined Docker Compose profiles that let you start small, scale fast, and run exactly what you need—no more, no less. Profiles align the platform with your intent, from first exploration to a full, production-grade agentic stack.
| Profile Name | Description |
|---|---|
try |
Minimal, zero‑friction setup to explore HyperAgency (default). Ideal for first‑time users, quick demos, and experiments. |
h9y |
Core HyperAgency services included in the h9y repository. Best for active development and local iteration. |
all |
Complete HyperAgency Cloud stack, mirroring the services available at HyperAgency Cloud. |
You can find the exhaustive list of available profiles in the docker-compose.yaml.
Choose the profile that matches your current goal by setting it in the .env file:
COMPOSE_PROFILES=h9yYou can enable multiple profiles and Docker Compose will merge them into a single runtime configuration:
COMPOSE_PROFILES=try,logs,metricsThen start HyperAgency using Docker Compose as usual. Docker will run only the services associated with the selected profiles.
Recommended for enterprise-grade deployments, high availability, and production-scale infrastructure.
Deploy HyperAgency on your own Kubernetes infrastructure using our official Helm chart.
- Full Kubernetes-native deployment
- Horizontal scaling of agents and services
- Production-grade observability integration
- Secure secrets management (Vault / external KMS)
- Multi-node distributed execution
📦 Helm chart and documentation provided upon request.
# Contact us to receive:
# - Helm chart (private or enterprise repo)
# - Cluster architecture guide
# - Production deployment instructions📩 Request access: artem@h9y.ai
Best for teams who want HyperAgency in production without managing infrastructure.
We deploy and operate a dedicated HyperAgency instance tailored to your environment.
- Dedicated cloud or hybrid deployment
- Custom configuration and integrations
- Continuous updates and maintenance
- Enterprise SLA (on request)
- Private tenant isolation
- Custom agents setup
- API integrations
- Observability dashboards
- Optional custom development
📩 Contact us for deployment: artem@h9y.ai
Fastest way to start using HyperAgency — no setup required.
Access HyperAgency directly via our hosted platform:
- Fully managed SaaS version
- Instant onboarding (no Docker / Kubernetes)
- Preconfigured agent ecosystem
- Web dashboard + chat interface
- Regular updates and new features
- Early experimentation
- Product evaluation
- Teams without DevOps resources
- Fast prototyping of agent workflows
👉 Create an account and choose a subscription plan: https://h9y.ai/signup
git clone git@github.com:vuics/h9y.git
cd h9y
docker-compose -f standalone.yml build
docker-compose -f standalone.yml upOpen localhost:3990 in the browser and start using standalone h9y.
git clone git@github.com:vuics/h9y.git
cd h9yCopy and customize .env files for the main platform and submodules:
cp env.example .env./gen-certs.shOn macOS, double-click each .crt file in ./certs/ to trust them in Keychain Access.
If your domain is h9y.localhost or anything on localhost, you may not need to configure the DNS, since the locahost often resolves to 127.0.0.1 automatically.
If your domain name does not resolve, set up /etc/hosts entries:
sudo ./setup-hosts.shUse Docker Compose to start all services:
docker-compose upOpen vault.h9y.localhost in the browser to unlock Vault. Vault is used for secure storage of keys, see the section "Initialize Vault" abovefor more details.
Open h9y.localhost in the browser. To unlock Vault for secure storage of keys, see the section "Initialize Vault" above.
docker-compose downThe HyperAgency source code is available under a multi-license model, allowing you to choose the license that best fits your use case.
| Use Case | License |
|---|---|
| Free for personal, educational, non-commercial use | Apache-2.0-NC |
| Free for commercial projects that remain open-source | AGPL-3.0 |
| Required for commercial closed-source, proprietary use, or SaaS offerings | Commercial License |
Answers on the most frequent licensing questions are listed in Licensing FAQ.
For commercial licensing inquiries, contact artem@h9y.ai.
HyperAgency isn’t just for experimentation — you can run real production-grade agentic systems in your environment under founder-led guidance.
Our 2–3 week pilot programs are designed to answer one critical question:
Can persistent, governed agents replace fragile AI automations in your organization?
- Live HyperAgency deployment in cloud or self-hosted environment
- 1–2 production-grade agentic systems solving a real operational problem
- Persistent memory, coordination, and observability running end-to-end
- Clear governance points with human approvals, audit logs, and permissions
- Concrete go / no-go decision for scaling agentic AI in your org
| Duration | Scope |
|---|---|
| 2 Weeks | Proof of viability with 1 agentic map (1–5 agents) |
| 3 Weeks | Operational readiness with 2 agentic maps (up to 10 agents + bridges) |
All pilots are hands-on, production deployments — not demos or slides.
Early participation lets you shape your autonomous workflows, get direct collaboration with the HyperAgency team, and jumpstart your agentic AI adoption.
HyperAgency is an evolving framework for building agentic, autonomous systems — and eventually, fully autonomous software companies.
It already offers powerful capabilities for orchestrating agents, coordinating distributed systems, and integrating LLMs and humans in real-time workflows. By joining early, you can contribute to defining what autonomous organizations of the future look like.
🔍 Explore. 🤝 Collaborate. 🚀 Build. Build your autonomous AI organization today — from fragile AI to living, compound intelligence systems.
- Website: h9y.ai
- Email: artem@h9y.ai
- LinkedIn: Artem Arakcheev
- Paid Pilot / Early Access: Apply here
