Effectiveness of a GPT-OSS-Based Automated Essay Scoring System and the Educational Value of LLM-Generated Feedback : A Case Study of Descriptive Writing Tasks | Synapse