Claude Skills by ViryaZheng

Session wrap-up: consolidates every descriptive markdown file (README.md, AGENTS.md, HANDOFF.md, docs/*) into ONE canonical CLAUDE.md, then reconciles it against the current code state so a fresh session or new teammate can pick up immediately. CLAUDE.md keeps a mandatory PROGRESS block at the top answering two questions: where are we now and what's the next concrete action. Trigger when the user says: "wrap", "wrap up", "/wrap", "tidy docs", "sync docs", "update docs", "archive", "this phase...

Forces discipline on every non-trivial request: restate the user's intent, diagnose the root cause, name the domain and reuse existing mature solutions before improvising, then deliver one chosen answer instead of a menu of options.

Lower the AI-detection score of English text (essays, papers, reports) while preserving every key term, number, citation, and the original logic. Uses a keyless ZeroGPT script as the objective baseline and the running agent itself as the rewriter — no API key. Reads Turnitin/GPTZero PDF reports for ground-truth scores. Use when the user wants to "降AI率 / 降低AI率 / 去AI味 / humanize AI text / bypass AI detector / lower Turnitin AI score / make this read human / reduce GPTZero score", or hands over ...