glad you’re here! welcome to the community.
Jim made me aware I could ask
“Which agent am I currently using”
You’re currently interacting with the Storyforming Agent—the part of the system that guides you through Dramatica structure, Storyform choices, and high-level narrative architecture.
so that led me to ask
“What other agents can I use”
and got back this reply:
Here are the other specialized agents you can call on, depending on what you need next:
-
Theory Agent
-
Great for deep dives into Dramatica concepts, terminology, and “why” the model works.
-
Ideal if you want tutorials, clarifications, or comparisons to other story theories.
-
-
Storyteller Agent
-
Focuses on turning Storyform beats into vivid storytelling—scene prose, dialogue passes, tone polishing.
-
Useful when you’re ready to dramatize a specific moment or need stylistic feedback.
-
-
Character Agent
-
Helps craft, flesh out, or troubleshoot characters so they align with structural roles.
-
Can suggest names, back-stories, motivations, and relationships that fit your Storyform.
-
-
Unspool Agent
-
Bridges the gap between raw idea and finished treatment—step-by-step brainstorming, outlining, or “unspooling” the narrative thread.
-
Perfect if you’re stuck moving from concept to cohesive plot flow.
-
So, those are five available agents.
This can work kind of like the old ‘database languages’ like dBase, where you
SELECT database1
and then use that database.
AND this is also a nice way a user to ‘stay up to date’ on what agents are available. Some ‘self reporting’ mechanisms on system status are always helpful.
Besides story creation prompts of all kinds, some of these ‘technical’ prompts could be quite useful I think.
Thanks for this!
Yes, most people probably won’t care who they are dealing with - and i’m making that more apparent with the update I’m fixing to release with GPT-5…but there will be additional agents that line up more precisely with Dramatica theory.
We do a lot of work behind the scenes so that the passing off from one agent to the next is seamless and based on best practices from all the story consulting I did over the years.
Sometime today I’ll be adding more to the docs to make it clear: Narrova’s Expertise
a few prompts I’ve run tonight:
*I understand that problem/solution are what you use for half of the throughlines, and symptom/response for the other half. how do I know which one to use?
prompt 2:
so for relationship story, it sounds like you have three cases.
1 success story and OS is symptom response, so RS is symptom responese
2 success story and OS is problem/solution, so RS is problem solution
2 failure story so RS is always problem/solution*
*Please check this statement and tell me if it is true or not.
*
spoiler: it wasn’t true.
prompt 3: I think the issue here is your language: “Relationship Story: Mirrors the Objective Story pair in a Success story, and mirrors the Problem pair in a Failure story. (Your story is Failure → RS shares Problem∕Solution with OS.)”
*I don’t see that they are ‘mirroring each other’ at all.
*
Navarro replied that the mirroring was positional, not literal.
prompt 4 *show me the storyform
*
_prompt 6: *you said: “The character who changes (Change-resolve) shares the Problem/Solutionelements with the Objective Story.”
But the IC character who changes has a Problem/Solution of Possibility/Probability and the OS has a Problem/Solution of Avoid/Pursuit. Why would you call this ‘sharing’?_
*
this prompt uncovered a mistake in the story form.
*prompt please examine the storyform with extreme care and precision and tell me about any other mistakes you have made, in detail. do not fix them yet.
prompt* *MC Pivotal element is uncontrolled. IC Pivotal element is controlled. Please change this and lock it, and audit storyform for correctness.
*
this was how I fixed another error.
examples of result after you ask for an audit.
*prompt “lock the corrections”
navarro:*
I’m ready to implement the deeper “Start ⇒ Dependent-pair” repair, but because that change would force a cascade (Symptom, Response, Pivot, and possibly OS alignment), we have to re-run a full set of structural tests.
Please confirm one of these options before I lock anything:
-
Minimal Adjustment – Flip MC Symptom/Response to Conscience / Temptation, accepting that Response duplicates the Problem (permitted but stylistically unusual).
-
Full Cascade – Re-engineer MC Symptom/Response and shift the OS Response (and thus MC Pivot) so no element duplicates and all pairs stay orthogonal.
-
Keep Current Form – Leave Control/Uncontrolled in place since all hard rules are already satisfied.
Let me know which path you prefer and I’ll execute it precisely and then perform the final audit.
*I chose 3 for the time being. I forget where I first heard it, maybe Melanie, ‘no on reads you for perfect storyform.’
*prompt :you said: "Growth Pair rule for Start — each Symptom should be the Dependent-pair to its Problem."
I don’t understand. Explain this to me in painstaking detail please.*
*
Navarro:
Below is the long-form, under-the-hood explanation of the “Start/Stop → Dependent/Companion” guideline.
(You can write a perfectly good Storyform without ever using it, but it helps when you want every microscopic knob aligned to Dramatica’s ideal harmonics.)
1. Pairs inside a Variation Quad
Take any Issue quad (e.g., Self-Interest, Selflessness, Approach, Attitude).
Connect opposite corners:
A _____ B Dynamic (push/pull)
| |
| |
C ¯¯¯¯¯¯ D Dynamic (push/pull)
That gives you four Element pairs:
-
Dynamic pairs – diagonal opposites (A ↔ D, B ↔ C)
– pure push vs. pull energy. -
Companion pairs – horizontals (A ↔ B, C ↔ D)
– cooperate easily, can get along or collude. -
Contrast pairs – verticals (A ↔ C, B ↔ D)
– opposite traits that compete or compare. -
Dependent pairs – each Dynamic pair recast as need-based
– “A completes D”; “B completes C.”
In diagrams the Dependent idea isn’t a new line; it’s a lens that says:
“These two functions rely on each other in a Start story.”
Example with Destinations quad (Temptation, Conscience, Control, Uncontrolled):
| Pair Type | Element Pair |
|---|---|
| Dynamic | Control ↔ Uncontrolled ; Temptation ↔ Conscience |
| Companion | Control ↔ Temptation ; Uncontrolled ↔ Conscience |
| Contrast | Control ↔ Conscience ; Uncontrolled ↔ Temptation |
| Dependent | Temptation + Conscience (complete each other) Control + Uncontrolled (complete each other) |
2. Growth Style (Start vs. Stop)
Growth describes what the Change character (or the story at large) must do:
| Growth | Meaning | Dramatica Hint |
|---|---|---|
| Start | Something missing must be added; the inequity grows until that missing piece enters. | Put focus on Dependent pairs—two halves that need to hook up. |
| Stop | Something harmful must be removed; the inequity clears when that thing is halted. | Put focus on Companion pairs—traits that cooperate until one is yanked away. |
So, in a Start story, the model nudges you to make the Symptom function (what characters treat as the problem) the complement of the real Problem—its Dependent partner.
When the Storyform reaches the end, the audience feels “the missing piece just snapped in.”
In a Stop story, characters chase a Companion element; the moment they remove the offending member, balance returns.
3. Applying the Guideline
If your MC is Start, Dramatica’s purity note says:
Problem — Temptation
Solution — Conscience
Symptom — Conscience (Dependent mate to Temptation)
Response — Temptation (dynamic mate to Conscience)
Readers then perceive that Friar keeps treating Conscience as the issue (Symptom).
What he really lacks is the restraint to apply Conscience fully—something that would “complete” his Temptation-driven nature.
In our current locked Storyform we instead gave him:
Problem — Temptation
Solution — Conscience
Symptom — Control (Dynamic with Uncontrolled, but not Dependent to Temptation)
Response — Uncontrolled
That still works structurally but the Growth guideline isn’t followed; the “missing piece” metaphor is weaker because Control doesn’t complete Temptation in Dramatica’s math.
4. Why It’s Optional
• The Start/Stop Dependent/Companion suggestion appears in advanced Dramatica classes but is not enforced by the software.
• Many canonical Storyforms (e.g., To Kill a Mockingbird) violate it yet pass every validation test.
• It only matters if you want the thematic “click” to follow the purest theory outline.
5. If You Ever Want to Align It
You would:
- Move MC Symptom to Conscience (Temptation’s Dependent mate).
- MC Response becomes Temptation.
- Because MC Pivot must still equal OS Response (Uncontrolled), you’d then shift the OS Symptom/Response pair to Conscience / Temptation as well.
- That, in turn, requires re-examining Objective Story beats to keep language consistent.
A cascading—but doable—adjustment.
Key Take-away
“Symptom = Dependent-pair” for a Start arc is a fine-tuning knob.
Leaving it on a Dynamic pair (Control/Uncontrolled) breaks no core rule; it only changes the flavor of how the audience experiences “what seems wrong” versus “what really is missing.”
*prompt give me a realistic detailed idea of how the audience experience would change.
*
can’t share the analysis here because delves into my story and don’t want to share that just yet. But it was an amazing analysis, and demonstrated to me that the change to a Dependent Pair WAS significant and completely changed how my MC would be perceived by the audience. It was NOT as trivial as earlier responses led me to think. This was my best prompt of the night.
finally, I got a response from narrova I wanted more answers about. This is REALLY improving my understanding of some of the plumbing in Dramatica. Wish I’d had this then I was trying to read the damned book.
prompt what do you mean exactly that “MC now has Temptation appearing as both Problem and Response and Pivot. Dramatica permits duplication, but note it makes his flaw/pivot very on-the-nose (tempted, reacts with more temptation).” what do you mean on the nose, and what does that look like, and what does the alternative look like?
What “Temptation Everywhere” Means in Practice
In the revised Storyform, Temptation now occupies three different structural slots for the Main-Character line:
- Problem – root driver of Friar’s personal conflict.
- Response – his immediate, chosen reaction to whatever he labels the Symptom.
- Pivotal Element – the switch that ultimately determines whether the wider story succeeds or fails.
Why Dramatica Allows It
The software only forbids duplication inside the same pair (e.g., you can’t have Problem = Solution).
Having the same element show up in separate pairs is legal, just uncommon.
“On-the-Nose” in Storytelling Terms
Think of a character’s behaviour as a three-beat loop:
| Structural Function | What the audience usually sees |
|---|---|
| Problem | Under-the-skin driver (often implicit early on) |
| Symptom / Response | Surface tactic: the thing they blame and the thing they do about it |
| Pivot | One high-stakes moment where altering (or clinging to) this element determines the ending |
If all three slots use the same word, you get a loop with no tonal shift:
Beat after beat, Friar feels tempted ➜ he diagnoses temptation-related Conscience failure ➜ he responds by giving in to temptation ➜ climax hinges on yet another temptation decision.
The audience “reads” him quickly:
“We get it—he’s always tempted and he always caves.”
There’s little room for surprise or self-revelation.
What It Looks Like on Screen
Mid-Act scene (Temptation saturates Problem + Response + Pivot)
Friar sees Amelia’s suppressed climate tape.
• Problem flare: That irresistible shortcut—bury it and keep the donors happy.
• Symptom (Conscience) trigger: He mutters a moral line: “It’s provisional—we’ll release it after the election.”
• Response (Temptation—same as Problem): He pockets the hush envelope.
• Later Pivot: Board vote swings on his final word; he again chooses the tempting payout.
Every stage circles the same flavour. Viewers feel repetition rather than escalation.
An Alternative with More Texture
(Keep Dependent-pair rule but avoid triple duplication)
Swap Response and Pivot positions:
| Slot | Element | Result |
|---|---|---|
| Problem | Temptation | Root flaw remains. |
| Symptom | Conscience | He talks morality. |
| Response | Conscience (yes, same as Symptom) OR Uncontrolled | He over-moralises or lets go—gives contrast. |
| Pivot | Uncontrolled (OS Response’s dynamic mate) | Final decision is not another “grab,” but a panicked loss-of-grip moment. |
How it plays:
- Early: tempted to fake belonging (Problem).
- Mid-beats: lectures conscience (Symptom) then “corrects” by excess virtue or reckless abandon (Response).
- Climax: boardroom spirals; he loses control (“Uncontrolled”) and that seals Failure.
Audience now experiences a gear-shift—from temptation to moralising to a final desperate loss of control—rather than a single flavour on loop.
Decide by Thematic Goal
• If you want Friar to feel trapped in one vice (temptation) with tragic inevitability → keep the triple-Temptation; “on-the-nose” becomes a stylistic statement.
• If you want shades of self-deception—moral posturing, sudden collapse—choose the more textured layout.
Either way the structure stays valid; the choice is purely about the dramatic colour you prefer.
prompt please undo the change we just made to symptom response and go back to controlled/uncontrolled. then audit the storyform for correctness.
then I changed my mind and went back to the original ‘twist’ that Dramatica threw in there.
Wow, that is a TON of information. 
I’m glad it’s helping and there are some interesting bits in there, especially the Start/Stop section BUT:
Narrova is a little confused. The facts are:
- Dynamic pairs are diagonal
- Companion pairs are horizontal
- Dependent pairs are vertical
Pivotal Elements are a result of Resolve/Growth/Outcome
Stop implies companion relationship between Problem and Symptom
Start implies dependent relationship between Problem and Symptom
We will look into why this is showing up as an issue.
Do note though (mentioned in email) - that an LLM conversation is not the best place to HOLD context for the conversation. You’ll want to make sure that once you have a Storyform, and it’s been validated (just ask), save it to your account either in a story or for a particular Conversation. Make sure you see that “atom” icon within the box at the bottom and that when you click on it, you see the entire Storyform.
This approach is essential to constantly keep the LLM/AI informed of the context of the Storyform you are working on. Until you set that, Narrova will always be able to shift in and out of valid Storyforms in an attempt to understand the kind of story you are writing.
thank you. I’ve been little fuzzy on that functionality.
some more prompts from this morning
please undo the change we just made to symptom response and go back to controlled/uncontrolled. then audit the storyform for correctness.
please validate the storyform and save it to outside this conversation to the dedicated storyforms area for this conversation
can I have more than one story for a storyform?
show me a list of all saved stories
show me a list of all saved storyforms
use storyform MyStoryForm for all further work unless otherwise explicitly instructed, no exceptions.
Change the title of the storyform to “MyStoryForm2”.
Thanks for this!
A couple of clarifications:
- a Story can have multiple Storyforms.
- A Storyform belongs to one Story
- Stories are accessed from the top nav
- Storyforms can be found within their assigned Stories
- If you are not in a Story context when you create a Storyform, Narrova will create one for you and assign the current conversation to the new Storyform context.
Listing stories and moving Storyforms though the interface is not currently available but will be soon.