LLM Platform Security: Applying a Systematic Evaluation Framework to OpenAI's ChatGPT Plugins