An open source platform for building, deploying, securing, and governing AI agents on Kubernetes.
Built on open standards:
- ecosystem-guide - Central reference for the Kagenti ecosystem
- kagenti - Main installer, UI, and documentation (Python)
- adk - Kagenti Agent Development Kit (Python)
- adk-starter - Starter template for building agents with the Kagenti ADK (Python)
- agent-examples - Sample agents and tools (Python)
- kagenti-operator - Kubernetes operator for deploying and lifecycle management of agents and tools (Go)
- kagenti-extensions - Kubernetes security extensions for zero-trust agent authentication (Go)
- agentic-control-plane - Control plane composed of specialized A2A agents coordinated through Kagenti CRDs (Python)
- mcp-gateway - Envoy-based MCP Gateway (Go) — hosted under Kuadrant org
- plugins-adapter - Adapter for security and safety plugins for Envoy-based MCP Gateway (Python)
- onecli - Open-source credential vault that gives your AI agents access to services without exposing keys (TypeScript)
- capture-the-flag - Capture the flag scenarios to demonstrate and test Kagenti's security posture
- workload-harness - Tools for agent load generation on the Kagenti platform (Python)
- .github - Project website and documentation (Hugo/HTML)
- 📖 Start with the CONTRIBUTING guide
- 👥 Meet the maintainers
- 🌐 Learn more at kagenti.io
- 💬 Or just come say hi in the Kagenti Discord