cjunekim/claude-plugins

Korean LLM-as-respondent toolkit built around the nvidia/Nemotron-Personas-Korea HuggingFace dataset (1M synthetic Korean adults). Bundles dataset-reference and dispatch-strategy skills, the domain-agnostic synthetic-population-validity skill (methodology + statistics for comparing synthetic/LLM survey distributions against a real human reference), the bulk-reply-save save pattern for batched dispatches, a persona-respondent sub-agent (stage-direction brackets + multi-item dispatch + file-write save mode), and a /persona-interviewee slash command for in-character qualitative interviews.