<2-3 sentences summarising what the project addresses and solves, the target audience, and the benefit.> kube-coder provides per-user, isolated development environments on Kubernetes, integrating cloud-native infrastructure with AI assistance. It targets developers and teams seeking rapid onboarding, persistent workspaces, and enterprise-grade security, revolutionizing productivity through zero-setup IDE pods.

How It Works

The system deploys isolated IDE pods per user via a Helm chart on Kubernetes. Each pod offers VS Code in the browser (code-server), persistent tmux terminals, and in-pod browser sessions. It leverages oauth2-proxy for GitHub OAuth authentication and integrates various AI assistants like Claude Code or OpenCode. This approach enables scalable, multi-tenant deployments with per-user namespaces, ingress rules, and persistent volumes for robust isolation.

Quick Start & Requirements

For local testing, prerequisites include Docker, minikube, kubectl, and helm, with a single command: make local. Cloud deployments require Kubernetes 1.19+, Helm 3.0+, an Nginx ingress controller, and a GitHub OAuth App. User onboarding involves make deploy-base followed by make new-user USER=<name> and make deploy USER=<name>. Detailed guides are available at docs/local-development.md and docs/NEW_USER_PROVISIONING.md. Video demos are linked in the repository.

Highlighted Details

• AI-Powered Workflows: Integrates Claude Code, OpenCode, Ante (DeepSeek), Gemini, and LibreFang, configurable via API keys and supporting persistent memory stores for context.
• Developer Productivity: Offers zero-setup onboarding, persistent workspaces that survive pod restarts, and multi-modal access including VS Code, terminal, and in-pod browser interfaces.
• Enterprise Security & Isolation: Enforces per-user isolation using separate Kubernetes namespaces, network policies, and persistent volumes, secured by GitHub OAuth and zero-trust networking principles.
• Operational Excellence: Leverages Kubernetes-native scaling and healing, Helm for infrastructure-as-code, and provides comprehensive monitoring metrics and health checks.

Maintenance & Community

The project maintains a subreddit at https://www.reddit.com/r/kubecoder/ for community interaction. Professional inquiries can be directed to scalebaseio@gmail.com. Specific contributor or sponsorship details are not prominent in the provided text.

Licensing & Compatibility

The project is released under the MIT license, which is permissive for commercial use and integration into closed-source applications.

Limitations & Caveats

Cloud deployment necessitates significant Kubernetes and OAuth configuration. Some AI assistants, like Claude, may incur usage costs. Components like the Ante CLI are in preview (0.preview.37), and LibreFang is in beta (v2026.6.10-beta.17), indicating potential for ongoing changes or instability in these specific features.