research-notebooklisted

# /research-notebook Command (NotebookLM Integration) ## Overview `*research-notebook` manages NotebookLM notebooks as persistent knowledge assets for TAD workflows. Use for research-intensive topics requiring cross-source synthesis (YouTube + PDF + web). **Key principle**: NotebookLM is a knowledge asset (stateful, persistent), not a stateless tool call. **Primary use: Alex design/discuss phase.** Also usable standalone (without /alex) via CLAUDE.md §2 research routing — see "Standalone Usage" below. 平台绑定交互决策（cross-harness binding）：本文件及其 references 中所有 AskUserQuestion 调用是「交互决策契约」而非具体工具——当前 harness 有该工具 （Claude Code）→ 直接调用；无该工具（Codex 等）→ 以编号纯文本列出全部选项 （1. … / 2. … / 3. …）并**停止等待用户输入**，用户以编号或自由文本作答； 禁止代答、禁止把选项折叠成默认值继续执行。SAFETY 门控的调用点（人工审批 / 归档确认 / 权限升级确认类）无论何种 harness 都必须获得真人作答后才能继续。 非交互执行模式（如 codex exec）→ 视为无人可答，按 blocked 停止并上报， 不得自选默认值；已按 YOLO/预授权模式运行且该决策点有书面预授权记录 → 按其协议处理，不适用本条 blocked 分支。 --- ## Standalone Usage (without /alex) When invoked via CLAUDE.md research routing (without /alex active): - Run preflight check (same as below) - Use the same CLI commands (create / ask / source add-research) - Skip Alex-specific protocols (Socratic, handoff, Gate, capability_pack_awareness) - After research completes → soft suggestion of next steps (see below) - When /alex IS active, Alex's own research protocols (research_notebook_awareness, research_plan_protocol) take precedence ### After Research Completes (Standalone) Inform user: "研究完成。Findings saved to {path}." Then sug