Features

KubeOpenCode brings Agentic AI capabilities into the Kubernetes ecosystem. Here's an overview of key features.

Core

• Live Agents - Persistent, always-running AI agents as Kubernetes services
• Flexible Context System - Provide knowledge to agents via Text, ConfigMap, Git, URL, and Runtime contexts
• Agent Configuration - Complete reference for all Agent spec fields
• Agent Templates - Reusable base configurations and ephemeral task blueprints
• Skills - Reusable AI agent capabilities from Git repositories
• Plugins - OpenCode plugins for deep agent customization
• Multi-AI Support - Use different agent images for various AI backends

Automation

• CronTask - Scheduled and recurring task execution
• Git Auto-Sync - Automatic sync with remote Git repositories
• Task Timeout - Automatic timeout for long-running tasks
• Task Stop - Stop running tasks via annotation
• Task Cleanup - Automatic cleanup of finished Tasks
• Task Session - OpenCode session info, token usage, and cost in Task status
• Concurrency & Quota - Limit concurrent tasks and rate of task starts

Collaboration

• Agent Share Link - Share terminal access via URL — no Kubernetes credentials required

Infrastructure

• Persistence & Lifecycle - Session/workspace persistence, suspend/resume, and standby
• Enterprise (Proxy, CA, Registry) - Corporate proxy, custom CA certificates, and private registry authentication
• Pod Configuration - Pod security, scheduling, system containers, and advanced settings