Argo cd

python/samples/argocd_install/.env_example

@@ -0,0 +1,7 @@
+model=gpt-4o-mini
+AZURE_API_KEY=<your-api-key>
+AZURE_API_BASE=https://<your-api-base>.openai.azure.com/
+AZURE_API_VERSION=2024-08-01-preview
+KUBECONFIG=/workspaces/autogen/python/samples/gitea_install/kubeconfig
+REPO_URL=https://github.com/<my-org>/<my-repo>
+REPO_PAT=<some-token-ending-with->11AVBLMCI0RB9l4J3GBnAC

python/samples/argocd_install/README.md

@@ -0,0 +1,40 @@
+# ArgoCD Install Example
+
+Agentic installation of ArgoCD on Kubernetes using local Kubectl/Helm CLIs. 
+
+## Prerequisites
+
+First, you need a shell with AutoGen core and required dependencies installed.
+- Use VSCode to start the DevContainer defined in `/.devcontainer/`.
+
+- Then install Python dependencies:
+```bash
+pip install "autogen-ext[openai,azure]"
+python3 -m pip install python-dotenv
+pip install autogen_agentchat playwright
+```
+
+## LLM Configuration
+
+The LLM configuration should be defined in a `.env` file. 
+Use `.env_example.yml` as a template.
+
+## Other Environment Variables
+
+In your `.env` file, also include the following: 
+- REPO_URL - Git repo used for storing ArgoCD App configs
+- REPO_PAT - Token (Fine-grained recommended) for accessing your Git repo at REPO_URL. 
+- KUBECONFIG - path to your kubeconfig file in the following section
+
+## Kubernetes Configuration
+
+The cluster configuration should be defined in a `kubeconfig` file. 
+Use `kubeconfig_example` as a template.
+
+## Running the example
+
+- Navigate to `/workspaces/autogen/python/samples/argocd_install/` and execute: 
+
+```bash
+python main.py
+```

python/samples/argocd_install/kubeconfig_example

@@ -0,0 +1,22 @@
+apiVersion: v1
+kind: Config
+clusters:
+  - name: rancher-desktop
+    cluster:
+      #server: https://127.0.0.1:6443
+      server: https://host.docker.internal:6443
+      certificate-authority-data: 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
+      insecure-skip-tls-verify: false
+users:
+  - name: rancher-desktop
+    user:
+      client-certificate-data: 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
+      client-key-data: LS0tLS1CRUdJTiBFQyBQUklWQVRFIEtFWS0tLS0tCk1IY0NBUUVFSUN3TzhYbnhUbnhsQjR6YVdDRGZYRk1YWG5YZzhTV0FHa0g5ZEpDTEUyZkpvQW9HQ0NxR1NNNDkKQXdFSG9VUURRZ0FFbnFZOWF3ZXM4MDBMcmRVWktpNVczamRMd0dzYkZuMm03VTczbWhhcnBpT3lxaFcvQ2xMdgpJOGpBR1E5NkUrNmo4NzdIbTVPVFFXbXplRUVIM2plbm5nPT0KLS0tLS1FTkQgRUMgUFJJVkFURSBLRVktLS0tLQo=
+contexts:
+  - name: rancher-desktop
+    context:
+      cluster: rancher-desktop
+      name: rancher-desktop
+      user: rancher-desktop
+preferences: {}
+current-context: rancher-desktop

python/samples/argocd_install/main.py

@@ -0,0 +1,111 @@
+from dotenv import load_dotenv
+import os
+import asyncio
+from pathlib import Path
+
+from autogen_agentchat.agents import AssistantAgent, UserProxyAgent, CodeExecutorAgent
+from autogen_agentchat.conditions import TextMentionTermination
+from autogen_agentchat.teams import RoundRobinGroupChat
+from autogen_agentchat.ui import Console
+from autogen_ext.models.openai import AzureOpenAIChatCompletionClient
+from autogen_ext.code_executors.local import LocalCommandLineCodeExecutor
+
+# Load environment variables
+load_dotenv("./.env")
+
+async def main() -> None:
+    """Main function to run Autogen agents and execute tasks sequentially."""
+
+    # Initialize Model Client
+    model_client = AzureOpenAIChatCompletionClient(
+        model=os.getenv("model"),
+        azure_endpoint=os.getenv("AZURE_API_BASE"),
+        api_key=os.getenv("AZURE_API_KEY"),
+        api_version=os.getenv("AZURE_API_VERSION"),
+    )
+
+    # Retrieve GitHub credentials from environment variables
+    repo_url = os.getenv("REPO_URL", "https://github.com/example/repo")  # Default if not set
+    repo_pat = os.getenv("REPO_PAT", "your-pat-token")
+
+    # Setup Code Execution Environment
+    work_dir = Path("coding")
+    work_dir.mkdir(exist_ok=True)
+    #local_executor = LocalCommandLineCodeExecutor(timeout=120, work_dir=work_dir)
+    local_executor = LocalCommandLineCodeExecutor(work_dir=work_dir)
+    code_executor_agent = CodeExecutorAgent("code_executor", code_executor=local_executor)
+
+    # Define Agents
+    assistant = AssistantAgent(
+        name="assistant",
+        model_client=model_client,
+        system_message="""Use the code_executor_agent to solve tasks. 
+            When writing code, provide at least one markdown-encoded code block to execute.
+            That is, quoting code in ```python or ```sh code blocks).
+            """,
+        reflect_on_tool_use=True,
+    )
+    user_proxy = UserProxyAgent("user_proxy")
+
+    # Termination condition
+    termination = TextMentionTermination("exit")  # Type 'exit' to end the conversation.
+
+    # Define Team
+    team = RoundRobinGroupChat(
+        [assistant, user_proxy, code_executor_agent],
+        termination_condition=termination
+    )
+
+    # Task 1: Install ArgoCD on Kubernetes using Helm
+    argocd_install_task = """
+    Install ArgoCD on Kubernetes using Helm, if not yet installed.
+    Install to the 'argocd' namespace.
+    Assume kubectl and Helm are pre-configured in the local environment.
+    Wait for ArgoCD pods to become responsive.
+    """
+    await Console(team.run_stream(task=argocd_install_task))
+
+
+    # Task 2: Ensure local access to ArgoCD server
+    argocd_access_task = """
+    Ensure local access to ArgoCD server
+    You may need port forwarding and / or a load balancer for ArgoCD server using kubectl. 
+    """
+    await Console(team.run_stream(task=argocd_access_task))
+
+
+    # Task 3: Install ArgoCD CLI and login to server
+    argocd_cli_install_task = """
+    Ensure the ArgoCD CLI is installed locally. Login to the server.
+    """
+    await Console(team.run_stream(task=argocd_cli_install_task))
+
+
+    # Task 4: Test repo access
+    repo_access_task = f"""
+    Use this GitHub repository: {repo_url}
+    Access the repository with this fine-grained token: {repo_pat}
+
+    Test repo read/write access.
+    """
+    await Console(team.run_stream(task=repo_access_task))
+
+
+    # Task 5: Configure ArgoCD to Deploy Argo Workflows via GitOps
+    argocd_config_task = f"""
+    Use ArgoCD's GitOps capabilities to deploy Argo Workflows (ArgoWF) to the argowf namespace, if not yet installed.
+    Store the ArgoWF config files in a GitHub repository for ArgoCD to scan. 
+    Store Helm values rather than simple K8s manifests. 
+
+    Retrieve the ArgoCD admin secret using:
+    ```sh
+    kubectl -n argocd get secret argocd-initial-admin-secret -o jsonpath="{{.data.password}}" | base64 -d
+    ```
+
+    Use this GitHub repository: {repo_url}
+    Access the repository with this fine-grained token: {repo_pat}
+    """
+    await Console(team.run_stream(task=argocd_config_task))
+
+# Run the script
+asyncio.run(main())