Implement call gpt script

Author: CuddlyBunion341

call_gpt.sh

@@ -0,0 +1,58 @@
+#!/bin/bash
+
+API_KEY=$OPENAI_API_RSMC_KEY
+MODEL="gpt-4o-mini"
+
+DOCS_DIR="docs"
+
+call_gpt_api() {
+  local prompt="$1"
+
+  response=$(curl -s -X POST https://api.openai.com/v1/completions \
+    -H "Content-Type: application/json" \
+    -H "Authorization: Bearer $API_KEY" \
+    -d '{
+      "model": "'"$MODEL"'",
+      "prompt": "'"$prompt"'",
+      "temperature": 0.7,
+      "max_tokens": 4096,  # Adjust this according to model limits
+      "top_p": 1,
+      "frequency_penalty": 0.2,
+      "presence_penalty": 0.2
+    }'
+  )
+
+  echo "$response" | jq -r '.choices[0].text'
+}
+
+for file in "$DOCS_DIR"/*/*.md; do
+  if [[ -f $file ]]; then
+    file_content=$(<"$file")
+    file_name=$(basename "$file")
+
+    prompt=$(cat <<PROMPT
+You are provided with a markdown document corresponding to the ECS plugin \'$file_name\'.
+
+1. Understand the document structure.
+2. Review the list of source files.
+3. Analyze the source code provided at the end to identify and describe (according to the template):
+   - The main **Components** and their purposes.
+   - The essential **Resources** with their roles.
+   - Various **Systems**, categorizing them appropriately, and descriptions of their functions.
+   - Networking aspects, if applicable.
+4. Create a corresponding mermaid diagram
+
+Ensure that you don\'t miss out on any details but that the documentation serves as a nice overview and is concise.
+Content:
+$file_content
+PROMPT
+)
+
+    response_text=$(call_gpt_api "$prompt")
+
+    output_file="${file%.*}_processed.md"
+    echo "$response_text" > "$output_file"
+    echo "Processed $file and saved the response to $output_file"
+  fi
+done
+