Docker Deployment
Run Hermes Agent in isolated, reproducible containers. Docker provides the ideal environment for safe agent execution with easy scaling and deployment.
Why Docker for AI Agents?
Docker containers provide isolated environments where AI agents can execute arbitrary code, install dependencies, and modify files without risking the host system. For autonomous agents that run shell commands and browser automation, containerization is a critical safety layer.
Quick Start
1. Pull the Image
docker pull nousresearch/hermes-agent:latest
2. Create Config
Mount a config directory with your API keys and preferences.
3. Run Container
docker run -v hermes-data:/data -v hermes-config:/config nousresearch/hermes-agent
Benefits
Isolation
Agent actions cannot affect the host system.
Reproducibility
Same environment everywhere — dev, staging, production.
Scalability
Spin up multiple agents with Docker Compose or Kubernetes.
Rollback
Revert to previous container snapshots instantly.
Related
Frequently Asked Questions
Is Docker required to run Hermes?
No. Hermes can run locally, on a VPS, via SSH, or in Docker. Docker is recommended for production because it provides isolation, reproducibility, and easy deployment.
What Docker image does Hermes use?
Hermes provides an official Docker image with Python, Playwright browsers, and all dependencies pre-installed. You can also build a custom image with additional tools.
Can Hermes run multiple containers?
Yes. You can run multiple Hermes instances in separate containers, each with its own configuration and memory. Docker Compose simplifies multi-instance orchestration.
How is persistent memory handled in Docker?
Hermes stores persistent memory in mounted volumes. The SQLite database and file storage live outside the container, ensuring data survives container restarts.