# Flokoa

## Docs

- [How Flokoa Works: Components and Architecture Guide](https://flokoa.dev/architecture.md): Understand how the Flokoa operator, Custom Resource Definitions, and runtime backends work together to manage AI agents on Kubernetes.
- [Give your agents access to external APIs with AgentTool](https://flokoa.dev/guides/add-tools.md): Create OpenAPI-backed AgentTool resources, attach them to your agents, and let the LLM call external APIs and internal Kubernetes services.
- [Connect a language model provider to your Flokoa agents](https://flokoa.dev/guides/connect-llm-provider.md): Create ModelProvider and Model resources to connect OpenAI, Anthropic, Google, or AWS Bedrock and attach a model to your running agents.
- [Deploy your first AI agent on Kubernetes with Flokoa](https://flokoa.dev/guides/deploy-first-agent.md): A step-by-step walkthrough for writing, applying, and verifying your first Flokoa Agent custom resource on a live Kubernetes cluster.
- [Run Flokoa agents with production-grade high availability](https://flokoa.dev/guides/high-availability.md): Configure multiple replicas, resource limits, liveness probes, and pod anti-affinity to keep your Flokoa agents resilient to failures.
- [Organize Flokoa resources across Kubernetes namespaces](https://flokoa.dev/guides/namespace-organization.md): Learn patterns for structuring Flokoa Agents, Models, ModelProviders, and AgentTools across namespaces to support teams and environments.
- [Flokoa: Kubernetes-Native AI Agent Platform for Teams](https://flokoa.dev/introduction.md): Flokoa is an open-source Kubernetes operator that manages AI agents as first-class CRDs — no custom runtime code required for template agents.
- [Install Flokoa on your Kubernetes cluster](https://flokoa.dev/operations/installation.md): Install the Flokoa operator and CRDs on any Kubernetes 1.25+ cluster using a single YAML bundle or the OCI Helm chart.
- [kubectl commands for managing Flokoa resources](https://flokoa.dev/operations/kubectl-reference.md): A complete reference of kubectl commands for creating, inspecting, updating, scaling, and debugging every Flokoa resource type.
- [Monitor and trace Flokoa agents with OpenTelemetry](https://flokoa.dev/operations/observability.md): Collect logs, expose metrics, configure distributed tracing with OpenTelemetry, and interpret agent status conditions in Flokoa.
- [Troubleshoot common Flokoa issues](https://flokoa.dev/operations/troubleshooting.md): Diagnose and resolve the most frequent problems with Flokoa agents, models, providers, tools, and the operator itself.
- [Get Started with Flokoa: Deploy Your First AI Agent](https://flokoa.dev/quickstart.md): Install the Flokoa operator, connect an OpenAI model provider, and deploy your first AI agent on Kubernetes in under five minutes.
- [Agent: deploy and manage AI agents on Kubernetes](https://flokoa.dev/resources/agent.md): Define, deploy, and manage AI agents as Kubernetes resources using the Agent CRD with declarative runtime, model, and tool configuration.
- [AgentTool: connect agents to external APIs](https://flokoa.dev/resources/agent-tool.md): Define OpenAPI-backed tool sets that your agents can call at runtime, targeting external URLs or internal Kubernetes services by name.
- [AgentWorkflow: orchestrate multi-agent pipelines](https://flokoa.dev/resources/agent-workflow.md): Define declarative multi-agent pipelines as Kubernetes resources that compile to Argo WorkflowTemplates for reliable, observable execution.
- [Instruction: manage agent system prompts as Kubernetes resources](https://flokoa.dev/resources/instruction.md): Store LLM system prompts as versioned Kubernetes resources, shareable across multiple agents, so prompts are auditable and easy to reuse.
- [Model: configure LLM models and parameters](https://flokoa.dev/resources/model.md): Declare a specific LLM model with generation parameters and a provider reference so your agents can consume it as a Kubernetes resource.
- [ModelProvider: connect Flokoa to LLM providers](https://flokoa.dev/resources/model-provider.md): Configure API credentials and connection settings for OpenAI, Anthropic, Google, and AWS Bedrock so agents across your cluster can access LLMs.
- [flokoa CLI: run and serve AI agents locally](https://flokoa.dev/sdk/cli.md): Use the flokoa CLI to start an A2A-compatible FastAPI server around any pydantic-ai or Google ADK agent, locally or inside a container.
- [Use Flokoa with Google ADK agents](https://flokoa.dev/sdk/google-adk.md): Wrap a Google ADK LlmAgent with Flokoa to serve it via the A2A protocol, inject operator-managed tools, and deploy it on Kubernetes.
- [Flokoa Python SDK: build and run AI agents](https://flokoa.dev/sdk/overview.md): The Flokoa Python SDK wraps pydantic-ai and Google ADK agents with an A2A protocol server, tool injection, tracing, and a CLI runner.
- [Use Flokoa with pydantic-ai agents](https://flokoa.dev/sdk/pydantic-ai.md): Wrap a pydantic-ai Agent with Flokoa to serve it via the A2A protocol, inject operator-managed tools, and deploy it on Kubernetes.