The Dramatica MCP plugin is now available in ChatGPT and Codex

[jhull]

The official Dramatica plugin is now available in ChatGPT and Codex!

Once connected, it brings the Dramatica Narrative Platform into the conversation you are already having. You can work in ordinary language while your assistant uses authenticated Dramatica tools instead of relying only on general model memory.

What is MCP?

MCP—the Model Context Protocol—is an open standard for connecting an AI assistant to outside tools and data.

In this case, MCP gives a compatible assistant a secure connection to Dramatica Theory, canonical example Storyforms, and the Stories and Storyforms in your authorized Dramatica account. The AI remains the interface; Dramatica supplies the structural intelligence and source material.

Because MCP is an open protocol, this is not limited to one interface. You can use the official Dramatica plugin in ChatGPT and Codex, or connect the Dramatica MCP server to another compatible client—including Claude—subject to that client’s plan and MCP support.

What can you do with it?

You can ask your assistant to:

• list your recent Dramatica Stories
• retrieve a saved Storyform or one specific Storypoint
• inspect the Storytelling and development context attached to your work
• validate a proposed Storyform without changing it
• compare your choices with canonical example Storyforms
• look up authoritative Dramatica definitions
• return to visible Dramatica conversations and their final answers
• propose an exact change, then save it only after you explicitly confirm it

For example:

List my five latest Dramatica Stories.

Get the Storyform for one of my saved Stories.

Validate this Storyform without changing it.

Whose AI usage does it use?

The AI model comes from the interface you choose.

If you connect Dramatica MCP to Claude and use Claude’s top model, Claude performs the reasoning and generates the response under your Claude plan. Dramatica MCP supplies the authenticated Dramatica data, tools, and validation; it does not make a second Narrova or OpenAI model response call behind the scenes.

So yes: an active Basic Dramatica subscription can provide access to Dramatica MCP while Claude provides the power of its top-tier models, subject to Claude’s current MCP availability and plan policies. Your AI client’s usage applies to the conversation; your Dramatica subscription governs access to Dramatica and its MCP tools.

For direct MCP clients, use:

https://platform.dramatica.com/mcp/dramatica

The simplest place to begin is a read-only request: ask Dramatica to define Story Judgment, list your latest Stories, or validate a proposed Storyform.

[SwiftCrisis]

Can I brainstorm with Claude about my projects using MCP in a manner similar to chatting with Narrova? I’m assuming there’s a lot going on behind the scenes with prompting and context to make Narrova give answers that fit Dramatica theory.

[jhull]

Great question.

Narrova Narrative Intelligence

There is, indeed, ALOT going on in-between the space when you ask Narrova a question, and when you get that response back. To say nothing of the mountain of research, best practices, and an as-yet-disclosed story intelligence system running in the background, it also persists your choices into discrete “objects” or components attached to your current Storyform.

The Development Ledger is kind of the running AI-notebook on what’s being decided, but when you really know parts of your story and you want to “stick a pin in them” - you can ask Narrova to save your Story Goal or your Main Character Problem, or any other aspect of your Storyform.

Narrova will oblige, and then you can later re-visit those in Subtxt – which was the original application we started building back in 2017. Subtxt is more of your classic “fill-in-the-boxes” application where you can see and remind yourself what you set in place for various aspects of your Storyform.

The two modules work very well together, and you can always bounce back-and-forth between Subtxt and Narrova as you develop. Again, this is up to the writer’s individual tastes – some like the free-form stream-of-consciousness dump that Narrova allows, while others want to see in concrete form all the various parts of their developing story.

The Dramatica MCP

The MCP (model context protocol) we setup is more about giving you writers the ability to access their work through the interface they will likely be using more and more going forward - which is to say almost all computer interaction will be through agents that do the nitty-gritty for you while you sit high above and steer things from afar. (Note this is about story development, not necessarily the actual writing process of telling a story).

The MCP gives your frontier agents access to many of Dramatica’s tools but definitely not all of them. You can pull up stories you’re working on, those individual components saved away I mentioned above, research storyform examples (a great way to learn Dramatica theory), or dive deep into certain aspects of Dramatica theory.

Here’s a good example of what you can and cannot do: let’s say you wanted to run some story coverage (analysis) on a recent screenplay draft. You wouldn’t really use the MCP because, as it stands now, it’s not running through the entire thought process of reading through the screenplay, deciding various aspects of your storyform, testing them against the Dramatica model, and then re-aligning that Storyforming process until it finds the most likely candidate. It’s only when you find that candidate (if you can), that you can then start to address issues with your story, what’s working and what’s not.

Once you have that analysis, and that Storyform, well then you could take that into Claude or ChatGPT and work on the iteration of it there. You would lose the Dramatica-focus on it, but maybe you want to pull away from it for awhile, or you feel more confident with some of the other tools you have and want to develop the story further there.

The MCP is there to give writers and story developers more capabilities as this technology grows and matures. You would come back to Dramatica and Narrova when you really want to have that objective partner, grounded in narrative theory, who won’t constantly tell you your story is amazing or come up with basic rote ideas that originate from captured storytelling, rather than conceptual storyforming.

That’s really the big difference between the two approaches.