edited the way api keys are handled.

Since API keys are now loaded from the auth server, this means that there is a delay between starting the software, loading the keys, and loading all the module files.
Due to nodejs running Async, the modules are being loaded BEFORE the api keys have been returned from the auth server.
So now the api keys are being set inside the module every time the module is being called.
This has absolutely no impact on performance, or security, plus, this was the easiest solution, especially as it required no big changes to the rest of the code

services/modules/llm-chat_gpt/chatgpt.js

@@ -8,7 +8,7 @@ if (!fs.existsSync(outputDir)) {
 }
 
 // Ensure SAIA API key is set in environment variables: export SAIA_API_KEY="your_api_key_here"
-const SAIA_API_KEY = process.env.SAIA_API_KEY; // Ensure SAIA API key is set in environment variables
+let SAIA_API_KEY // Ensure SAIA API key is set in environment variables
 const SAIA_URL = "https://chat-ai.academiccloud.de/v1/chat/completions"; // URL for the REST call, used model and action
 
 const module_exports = {
@@ -18,6 +18,7 @@ const module_exports = {
     description: "Generates documents using OpenAI GPT OSS 120B via SAIA platform",
 
     async function(parameter) {
+        SAIA_API_KEY = process.env.SAIA_API_KEY;
         return new Promise(async (resolve, reject) => {
             try {
                 // console.log("SAIA OpenAI GPT module invoked with parameters:", parameter);