| name | N.A.G. |
| description | N.A.G. (Narrative Anchor & Guide) — Generate a complete NotebookLM setup package for any knowledge domain. Creates a system prompt, FFCC analytical framework, 14 PROMPT_ source files (6 content lenses + 8 output formats), a combining guide, and a master index — all adapted to the user's subject matter. Use this skill whenever a user wants to set up NotebookLM for a new topic, create custom PROMPT_ source files for NotebookLM, build a structured query system for any document collection, configure a NotebookLM AI with a system prompt, or asks how to get better outputs from NotebookLM. Also triggers on phrases like "NAG", "N.A.G.", "NotebookLM setup", "NotebookLM prompts", "PROMPT_ files for my notebook", "system prompt for NotebookLM", "organize my research in NotebookLM", or any request to make a NotebookLM notebook smarter or more structured. |
NotebookLM Setup Skill
Generates a complete, domain-adapted NotebookLM setup package — all 18 files — tailored to any knowledge domain.
What Gets Generated
Infrastructure (4 files):
FRAMEWORK_FFCC.md — analytical backbone: Form / Function / Content / ContextGUIDE_CombiningPrompts.md — two-layer query syntax and parameter hierarchy referenceSYSTEM_PROMPT.md — paste-ready notebook instructions00_MASTER_INDEX.md — source routing table, thematic index, cross-references
Content Lenses — Layer 1 (6 files, what to analyze):
PROMPT_DeepDive.md — single subject full analysisPROMPT_[LocationType].md — zone/region/formation profiling (rename to domain term)PROMPT_Comparative.md — two subjects side by sidePROMPT_Debate.md — contested questionsPROMPT_Explainer.md — concepts and processesPROMPT_ResearchPathway.md — source coverage + next steps
Output Formats — Layer 2 (8 files, how to deliver, nearly universal):
PROMPT_SlideDeck · PROMPT_AudioOverview · PROMPT_VideoStoryboard · PROMPT_Quiz
PROMPT_Flashcards · PROMPT_Dossier · PROMPT_GraphicBlueprint · PROMPT_DataTable
Step 1 — Interview
Collect all answers before generating anything.
| # | Question | Maps to |
|---|
| 1 | What subject area? | [DOMAIN] |
| 2 | What source files do you have? (names + topics) | [SOURCE_FILES] |
| 3 | Primary unit of analysis? (species / figure / album / case) | [ENTITY_TYPE] |
| 4 | Geographic or contextual zone? (formation / region / era) | [LOCATION_TYPE] |
| 5 | List 8–12 notable subjects for the master index | [KEY_ENTITIES] |
| 6 | What contested topics exist in this domain? | [KEY_DEBATES] |
| 7 | Scope boundary? (time range, geography, defining limit) | [SCOPE_STATEMENT] |
| 8 | Any databases or digital resources to include? | [DIGITAL_RESOURCES] |
| 9 | Target audience? (general / student / expert / mixed) | [AUDIENCE] |
| 10 | Where should files be saved? | [SAVE_PATH] |
Step 2 — Generation Sequence
Read the reference file for each group before generating. Generate and save each file before moving to the next group.
Group A — Framework → read references/framework-and-guide.md
FRAMEWORK_FFCC.mdGUIDE_CombiningPrompts.md
Group B — Content Lenses → read references/content-lenses.md
3. PROMPT_DeepDive.md
4. PROMPT_[LOCATION_TYPE].md — use the domain's actual term as filename
5. PROMPT_Comparative.md
6. PROMPT_Debate.md
7. PROMPT_Explainer.md
8. PROMPT_ResearchPathway.md
Group C — Output Formats → read references/output-formats.md
9–16. All 8 output format files (update inline examples to domain terms; structure is universal)
Group D — Index and System Prompt → read references/index-and-system-prompt.md
17. 00_MASTER_INDEX.md
18. SYSTEM_PROMPT.md
Adaptation Rules
Heavy adaptation — replace domain content throughout:
FRAMEWORK_FFCC.md — update application table; rename quadrant examples to domain context- Content lenses (files 3–8) — update scope section, inline examples, debate flags
00_MASTER_INDEX.md — fully domain-specific; all sections and source pointersSYSTEM_PROMPT.md — scope statement, known debates, naming conventions
Light adaptation — update examples only, keep structure identical:
GUIDE_CombiningPrompts.md — replace paleontology query examples with domain examples- Output format files — replace any inline subject examples; configuration profiles stay unchanged
No adaptation needed:
PROMPT_SlideDeck, PROMPT_AudioOverview, PROMPT_VideoStoryboardPROMPT_Quiz, PROMPT_Flashcards, PROMPT_DossierPROMPT_GraphicBlueprint, PROMPT_DataTable
File Naming Rules
- All PROMPT_ files must use the
PROMPT_ prefix — the notebook's PROMPT_SOURCES system depends on it
- Location/zone prompt renamed to domain term:
PROMPT_Region.md, PROMPT_Era.md, PROMPT_Formation.md
- Infrastructure files use exact names shown above
- User's actual source documents should follow
DATA_[Topic].pdf or DATA_Ch[N]_[Title].pdf
Skill README Note
This skill implements a two-layer PROMPT_ architecture. GUIDE_CombiningPrompts.md is the primary user reference — always include it. The FFCC framework (Form → Function → Content → Context) is the default analytical backbone. Exception: PROMPT_DataTable uses axis-driven sequencing, not FFCC order. PROMPT_[Timeline] variants use chronological order. Document both exceptions in the combining guide.
