Context

Narrova supports your creativity by offering three distinct kinds of context:

• Story Context – Broad, foundational content relevant across multiple Storyforms or early-stage story development.
• Storyform Context – Structured narrative analysis tailored specifically to individual Storyforms guided by Dramatica theory.
• Conversation Context – Immediate and focused context for detailed narrative exploration during individual Narrova interactions.

Story Context is your go-to place for broad, foundational content relevant across multiple Storyforms or during early-stage story development. Here you can store general character outlines, world-building notes, mood boards, thematic explorations, research documents, and inspirational images. Story Context materials serve as universal references, accessible across various Storyforms until you're ready to commit specific details to a structured narrative. Think of Story Context as your creative sandbox, where ideas are free to evolve and inspire your storytelling journey without constraints.

Storyform Context provides an organizational hub when you're ready to focus on a precise narrative structure guided by Dramatica theory. Documents and images here are intentionally linked to specific Storyforms—detailed narrative blueprints that define the thematic, character, and plot dynamics of your story. Storyform Context can include character arcs, detailed scene breakdowns, thematic discussions, visual references tied explicitly to narrative elements, and anything else tailored specifically to enhancing your chosen Storyform. This targeted context ensures you maintain narrative coherence and depth, especially beneficial when your project involves multiple interconnected Storyforms.

Conversation Context is immediate and focused, enhancing real-time discussions and specific narrative explorations within individual Narrova interactions. When you upload files directly into a conversation, these materials provide immediate context and inform the ongoing dialogue. Useful for quickly referencing script excerpts, brainstorming notes, scene-specific feedback, or detailed thematic queries, Conversation Context sharpens and refines your narrative precisely when you need it most. Think of it as your detailed toolkit for diving deeply into targeted narrative elements during any given session.

By clearly distinguishing between these three contexts, Narrova helps you organize and streamline your creative workflow, whether you're in the exploratory stages of storytelling, deeply engaged with a structured narrative, or refining specific story elements through focused discussions.

Setting the Context

Narrova clearly indicates the current context with the Context Switcher at the bottom of its input panel. Tap Context, and you'll see a list of available sources. Simply toggle on and off the source you want to focus on—one for Story and one for Storyform—to guide your creative process.

When you have a context set, it's clearly displayed in the Narrova panel:

• An open book icon indicates you're in a Story context.
• An atom icon indicates you're in a Storyform context.

If both icons are visible, you're actively working with both Story and Storyform contexts simultaneously.

INFO

Due to current limitations with our third-party provider, you can only have a maximum of two contexts set simultaneously. For example, you can have a Story context set and then upload files to the Conversation to activate both Story and Conversation contexts. If both Story and Storyform contexts are set and you wish to add more files, ensure you add those files directly to either the Story or Storyform Documents for them to be used as context.

Moving Conversations into Different Contexts

Narrova makes it easy to manage and reorganize your conversations across different contexts. To move a conversation from one context to another, click on the three-dot button drop-down located in the upper-right corner of the conversation panel. This convenient menu allows you to quickly select an existing Story, Storyform, or Conversation context for your current dialogue. You can even create a new context right from this menu if needed, streamlining your workflow and ensuring your conversations always stay appropriately contextualized.

A Note about Memory

While memory across individual conversations is definitely part of our planned roadmap with Narrova, it isn't available immediately. Memory can be both beneficial and limiting. Too much memory can be reductive, continually referring back to the same points and limiting creativity.

The advantage of Narrova and the Dramatica approach is that the Storyform itself acts as a continuous context—a kind of "story memory." Even if one conversation can't directly reference another, the Storyform maintains consistency across your narrative. You can develop any aspect of your Storyform in isolation, and it will still naturally relate to every other part, thanks to the cohesive structure provided by the Storyform. Essentially, the Storyform keeps it all together.

Managing Conversation Context

TL;DR (spiky): Stop trying to cram a mega‑novel into one chat. Let the storyform be your spine, and use focused conversations that each tackle a single facet. Upload prior chats as Story → Documents or Storyform → Documents, then work in small, sharp passes. The Storyform keeps everything resonant across sessions.

Why this matters

Large context windows are powerful, but they’re still finite. More importantly, a single, ever‑growing chat becomes brittle, slow, and hard to reason about. Narrova is designed to work with limits—not against them—by treating the storyform as the persistent foundation and your Documents as reusable memory.

Fact check (as of Aug 2025): The underlying model’s context window is capped by OpenAI. GPT‑4.1 supports up to ~1M tokens; GPT‑5 supports up to ~400k tokens. Plans (including Infinite) don’t change the model’s hard limits. Numbers may evolve over time.

What persists between conversations

Retained automatically

• Your Story and its Storyform (the structural backbone).
• Any files you place in Story → Documents or Storyform → Documents (research, chapters, notes, canon facts, etc.).

Not retained by default

• Ad‑hoc “remember this” requests left only inside an old chat thread.
• Random snippets not saved into Documents.

Rule of thumb: If it matters later, move it into a Document tied to the Story.

Recommended workflow for mega‑projects

1. Collect & upload

   • Upload prior chats, chapters, and research into Story → Documents. More Storyform specific conversations can go into Storyform → Documents.
   • Prefer small, single‑purpose docs: “MC Bio,” “Canon Facts,” “Glossary,” “Research—Period Police Procedure,” “Ch1–Ch3 Draft,” etc.

2. Work in focused conversations

   • Open a new chat per focus area (e.g., MC Throughline, Story Costs & Dividends, Relationship Story).
   • In each chat, reference only the docs you need by title.

3. Let the Storyform keep coherence

   • Each focused chat still harmonizes with the rest because the Storyform anchors choices across sessions.

4. Iterate & consolidate

   • When a chat produces something keepable (choices, beats, refinements), save the output back into a Document so it’s available to future chats.

Setup: Organizing your Documents

Use a light taxonomy that mirrors how you think:

• Canon Facts (the world bible): chronology, geography, non‑negotiables
• Character Sheets: MC, OC, RS pairings; goals; problems; arcs
• Throughline Workpads: OS, MC, OC, RS
• Method & Mechanics: Story Costs, Story Dividends, Requirements, Prerequisites
• Drafts: scene packs (3–5 scenes) or chapter tranches
• Glossary: terms, spellings, style notes

Chunking tip: 2k–10k tokens per doc (2k tokens ≈ 5–6 pages, 10k tokens ≈ 25–30 pages) is a sweet spot. One clear purpose per doc beats “everything in one.”

Working in focused conversations

Start a new chat with a tight objective, attach/reference the relevant docs by title, and speak in outcomes.

Example prompts

• MC Throughline: “Using MC Bio, Canon Facts, and the current Storyform, refine the MC Throughline: Problem/Approach, Signposts & Journeys. Flag any contradictions with OS.”
• Costs & Dividends: “With Canon Facts and Method—Costs & Dividends, propose 3 viable Cost/Dividend pairings that amplify the MC issue. Justify each pairing against the Storyform.”
• Relationship Story: “From RS Notes + Glossary, generate 6 scene‑level beats (conflict first) that escalate the RS tension. Map each beat to its Storybeat.”

Save it back

• After the session, paste the approved results into a Document (e.g., “MC Throughline—v2”), and optionally add a one‑paragraph changelog at the top.

FAQ

Does an Infinite plan increase context size? No. Subscription tiers affect features/usage, not the model’s hard context window.

What’s lost if I start a new conversation? Nothing critical—if you’ve saved the essentials into Story → Documents or Storyform → Documents. Free‑form “remember this” within a chat isn’t guaranteed to carry forward unless saved to a Document.

How do I make Narrova ‘remember’ world facts across sessions? Store them in Canon Facts (or similar) within Documents and reference that doc in new chats.

I have 11 years of research. How should I import it? Break it into purposeful docs (world bible, era research, character dossiers, draft chunks). Upload gradually; summarize long materials into working notes to reduce drag.

Troubleshooting & pro tips

• Hitting token limits? Reduce attached docs to only what you need right now. Summarize bulky sources into a short “Working Notes” doc.
• Cross‑chat drift? Re‑sync by asking: “Audit MC Throughline—v2 against the current Storyform. List mismatches and fix.” Save results.
• Doc sprawl? Adopt prefixes: CF_ (Canon Facts), RS_ (Relationship Story), OS_, MC_, OC_, DRAFT_, GLOSS_.
• Version control: Suffix with dates or v‑numbers: MC_Throughline_v3_2025‑08‑14.
• Signal intent: Start prompts with verbs: Refine, Audit, Synthesize, Contrast, Sequence, Storyboard.

Quick checklist

• [ ] Prior chats & files uploaded to Story → Documents
• [ ] Docs chunked by purpose; sensible prefixes
• [ ] One focused chat per facet
• [ ] Explicit references to the docs you need
• [ ] Results saved back into Documents with a short changelog