From notebooklm-enhanced
Multi-notebook research synthesis — activates when user needs cross-notebook research, comparative analysis, or deep research workflows combining multiple NotebookLM notebooks
npx claudepluginhub fakoli/fakoli-plugins --plugin notebooklm-enhancedThis skill uses the workspace's default tool permissions.
Multi-notebook research synthesis: query across notebooks, compare findings, and produce unified analyses.
Compares coding agents like Claude Code and Aider on custom YAML-defined codebase tasks using git worktrees, measuring pass rate, cost, time, and consistency.
Designs and optimizes AI agent action spaces, tool definitions, observation formats, error recovery, and context for higher task completion rates.
Designs, implements, and audits WCAG 2.2 AA accessible UIs for Web (ARIA/HTML5), iOS (SwiftUI traits), and Android (Compose semantics). Audits code for compliance gaps.
Multi-notebook research synthesis: query across notebooks, compare findings, and produce unified analyses.
Explicit triggers:
Intent detection:
/notebooklm-enhanced:setup)All commands use:
uv run --project "${CLAUDE_PLUGIN_ROOT}/scripts" notebooklm ...
When the user wants to compare or synthesize across existing notebooks:
List all notebooks:
uv run --project "${CLAUDE_PLUGIN_ROOT}/scripts" notebooklm list --json
Identify relevant notebooks based on the user's topic. Look at titles and source counts.
Query each notebook individually:
# Set context to each notebook and query
uv run --project "${CLAUDE_PLUGIN_ROOT}/scripts" notebooklm use NOTEBOOK_ID_1
uv run --project "${CLAUDE_PLUGIN_ROOT}/scripts" notebooklm ask "RESEARCH QUESTION" --json
uv run --project "${CLAUDE_PLUGIN_ROOT}/scripts" notebooklm use NOTEBOOK_ID_2
uv run --project "${CLAUDE_PLUGIN_ROOT}/scripts" notebooklm ask "RESEARCH QUESTION" --json
Synthesize findings: Combine answers from all notebooks, noting which notebook each finding came from. Identify:
Present the synthesis in a structured format with attribution to source notebooks.
When the user wants comprehensive research on a topic they haven't explored yet:
Create a dedicated research notebook:
uv run --project "${CLAUDE_PLUGIN_ROOT}/scripts" notebooklm create "Research: TOPIC" --json
uv run --project "${CLAUDE_PLUGIN_ROOT}/scripts" notebooklm use NOTEBOOK_ID
Add web research sources:
uv run --project "${CLAUDE_PLUGIN_ROOT}/scripts" notebooklm source add-research "TOPIC" --mode deep --no-wait
Wait for research completion:
uv run --project "${CLAUDE_PLUGIN_ROOT}/scripts" notebooklm research wait --import-all
Run synthesis queries:
uv run --project "${CLAUDE_PLUGIN_ROOT}/scripts" notebooklm ask "Comprehensive overview of TOPIC" --json
uv run --project "${CLAUDE_PLUGIN_ROOT}/scripts" notebooklm ask "What are the key debates and open questions?" --json
Generate a report artifact:
uv run --project "${CLAUDE_PLUGIN_ROOT}/scripts" notebooklm generate report --format briefing-doc --json
Present findings to the user with key takeaways.
For fully autonomous execution of complex research workflows, delegate to the research agent:
Use the Task tool to launch the
research-agentwith the user's research request. The agent handles the full pipeline (create, source, wait, query, generate, download) without user intervention.
Present research results with:
/notebooklm-enhanced:research — Single-step research workflow/notebooklm-enhanced:query — Query a single notebook/notebooklm-enhanced:library — Browse all notebooks/notebooklm-enhanced:generate — Generate artifacts from a notebook