Anonymized dataset of FGCC interactions

246,985 FGCC events from 136 computer science students engaged in controlled programming tasks. Events capture 23 IDE interaction types including character edits, bulk edits of source code, test results, user prompts to LLM and and LLM responses. The original data was anonymized - data in the 'file_name', 'old_text', 'new_text', 'surrounding_text', 'additional_data', and 'source_file' columns was replaced with random strings of characters of length corresponding to the original data. The time dependencies and assignment of datapoints to FGCC operation types, as well as the assignment of the user generating the event, remained unchanged.