log_analysis_multi_agent_rag: Support API key environment variables from main documentation

community/log_analysis_multi_agent_rag/README.md

@@ -17,7 +17,7 @@ This repository provides a sample code to demonstrate how you can use the log an
 - **Generate your API key** by following the steps in the link below:
   [Click here to view the steps for generating an API Key](https://docs.nvidia.com/nim/large-language-models/latest/getting-started.html#generate-an-api-key)
 
-- **Store your API key** : You can securely store your API key by creating a `.env` file in the root directory of your project
+- **Configure your API key** : Following the [examples in the document above](https://docs.nvidia.com/nim/large-language-models/latest/getting-started.html#export-the-api-key), place your API key in your shell initialization files. Alternatively, you can use environment variable managers like [direnv](https://direnv.net/).
 - **example.py** : The sample script showcases how to integrate log analysis into your workflow. It demonstrates how to pass your log data through the system, generate insights, and manage the output.
 
 # Components

community/log_analysis_multi_agent_rag/utils.py

@@ -65,8 +65,16 @@ def format_docs(self, docs):
 # Usage
 
 
-# Access the API key from environment variables
-api_key = os.getenv('API_KEY')
+# Access the API key from environment variables preferring the well known environemnt variables.
+# Finally falling back to the .env environment variable within this repo
+api_key = (
+    os.getenv('NVIDIA_API_KEY') or
+    os.getenv('NGC_API_KEY') or
+    os.getenv('API_KEY')
+)
+if not api_key:
+    raise RuntimeError("No NVIDIA API key found. Set one of NVIDIA_API_KEY, or NGC_API_KEY, or .env based API_KEY in your environment.")
+
 model = "nvidia/llama-3.3-nemotron-super-49b-v1.5"
 prompts_file = "prompt.json"
 automation = Nodeoutputs(api_key, model, prompts_file)