The Pacing Autopsy: AI Prompts That Diagnose Slow Chapters Before Your Editor Does
Why Pacing Fails Silently
Most novelists know within a few pages when something feels off about a chapter. The prose is clean, the dialogue sounds right, the character work is present—and yet the chapter sits there like a stone. Readers would drift. You can feel it even if you can't name it.
The harder problem is that slow chapters and quiet chapters look nearly identical on the page. A scene where your protagonist sits alone and processes grief can be essential structural work—slowing the reader down before a catastrophic second-act turn. Or it can be 1,400 words of narrative wheel-spinning dressed up as interiority. The difference isn't word count or action level. It's whether the scene is doing purposeful work toward the book's forward motion.
This is exactly where AI assistance becomes complicated. If you paste a chapter and simply ask "is this too slow?", you'll get a hedge. The AI doesn't know your intention. It doesn't know whether the stillness is deliberate. It doesn't know whether the chapter's job is to accelerate tension or metabolize grief or establish a relationship dynamic before everything breaks apart.
That means your first task before any diagnostic prompting is defining intent. What is this chapter supposed to accomplish? What should a reader feel or understand by the final paragraph that they didn't at the opening? Once you can articulate that in a sentence or two, you've given the AI the standard against which to measure the chapter's actual performance. Without it, you're asking for an opinion. With it, you're requesting an audit.
The Four Diagnostic Categories
Before running any AI workflow on a chapter, it helps to understand the four most common pacing failure modes. Each has a distinct signature, and each requires a different fix. Conflating them leads to revision that addresses symptoms instead of causes.
Scene Compression Problems
This is the "too much happening too slowly" failure. The chapter contains multiple meaningful events—an argument, a revelation, a decision—but each is padded with setup, transition, and re-explanation until the cumulative weight buries the momentum. Individual scenes aren't boring; they're just overlong relative to their narrative weight.
Interiority Bloat
Character thought-loops that don't resolve. The protagonist thinks about the same fear, the same conflict, the same unresolved relationship across three separate paragraphs, each circling without landing anywhere new. This is one of the most common failures in literary and character-driven fiction because it feels like depth when you're writing it and like stalling when you're reading it.
Beat Redundancy
Scenes that restate what has already been established rather than advance anything. A character learns something important in chapter seven. In chapter eight, they reflect on it. In chapter nine, they have a conversation that re-explains it to another character. The information is solid; the chapter's job is to move from that information toward consequence. Beat redundancy happens when writers are uncertain the reader absorbed the earlier scene and re-teach it instead of building on it.
Tension Flatline
No rising or falling stakes within the chapter itself. Every scene should have some internal arc—pressure that builds or releases, even minimally. A flatline chapter maintains the same emotional register from first paragraph to last. Even quiet chapters need micro-tension: the threat of what might be said, the avoidance of a topic, the thing someone almost does. Without it, readers have no propulsive reason to continue.
Prompt Workflow Step 1 — The Pacing Triage
The first diagnostic prompt is designed to give you a momentum map of your chapter. Rather than asking the AI to judge the whole thing at once, you're asking it to rate forward motion in discrete blocks—roughly every five paragraphs—and identify where that motion stops and why.
Before you paste the chapter, write a one-to-three sentence statement of intent. Something like: "This chapter is meant to establish that Marcus has decided to leave but hasn't told anyone yet. The emotional temperature should be restrained dread. The reader should end the chapter more anxious than they began it." Paste that before the chapter text.
I'm going to share a chapter from my novel along with a brief statement of its intended purpose. I need you to perform a pacing audit, not a general critique. Chapter intent: [paste your 1–3 sentence statement of what this chapter is supposed to accomplish emotionally and narratively] Here is the chapter: [paste full chapter text] Please divide the chapter into five-paragraph blocks and rate the narrative momentum of each block on a scale of 1–5, where 1 means forward motion has stopped entirely and 5 means the scene is pulling the reader actively forward. For each block, write one sentence explaining the momentum rating and one sentence identifying the primary cause if the rating falls below 3. After the block-by-block breakdown, give me a single paragraph identifying the chapter's lowest-momentum stretch and categorizing its failure type as one of the following: scene compression, interiority bloat, beat redundancy, or tension flatline. If more than one failure type is present, rank them by severity. Do not rewrite any prose. Do not suggest the chapter is good or bad overall. I need a diagnostic, not a verdict.
What you'll get back is a momentum map—a block-by-block picture of where the chapter is moving and where it's stalled, with a primary failure category attached. That category is what you take into the second round of prompts.
Prompt Workflow Step 2 — The Surgical Fix Prompts
Once you have a failure category from the triage, you move to targeted repair prompts. Each is designed for a specific diagnosis and asks the AI to suggest structural interventions without rewriting the chapter wholesale. The goal is a revision strategy you can execute yourself, not a replacement draft.
For Interiority Bloat
This prompt asks the AI to locate thought-loops and propose compression or redirection—either cutting the loop entirely, landing it in a single sharper sentence, or converting internal thought into an external beat that reveals the same information through action or dialogue.
Here is a passage from my chapter that has been flagged for interiority bloat—the character's internal thought-loops are circling without advancing: [paste the flagged passage] The character's name is The emotional situation is: [brief description—e.g., "she is avoiding the realization that she caused the accident"]. The chapter's job in this stretch is to show [intended effect—e.g., "that she cannot yet accept what she knows to be true"]. Please do the following: 1. Identify every distinct thought-loop in the passage—note where the same emotional territory is covered more than once. 2. For each loop, suggest one of three options: (a) cut it entirely if it adds nothing new, (b) compress it to a single sentence that accomplishes the same work, or (c) convert it to an external beat—a physical action, a gesture, a piece of dialogue—that shows the same emotional state without narrating it. 3. Indicate which option you'd recommend for each loop and why. Do not rewrite the passage. Provide the strategy; I'll execute the revision.
For Beat Redundancy and Tension Flatline
This prompt works on structural sequence—either collapsing redundant beats or injecting micro-conflict into a flatline section. It's particularly useful when a chapter's second half loses altitude after a strong opening.
I'm working on a chapter passage that has been diagnosed with [beat redundancy / tension flatline—choose one]. Here is the passage: [paste the specific stretch] Context: The chapter comes after [brief summary of what immediately precedes it in the narrative]. The information or emotional state this passage is covering was already established in [brief description of earlier scene or chapter]. The chapter needs to end with the reader feeling [intended emotional state—e.g., "that a confrontation is unavoidable"]. If the diagnosis is beat redundancy: Identify which specific beats are restating established information rather than advancing to consequence. For each redundant beat, suggest what consequence, complication, or new information could replace it while preserving any character work the beat is doing. Do not suggest cutting the character work—only cutting the restatement. If the diagnosis is tension flatline: Identify three moments in this passage where a micro-conflict could be inserted without distorting the scene's tone. A micro-conflict can be as small as an avoidance, an almost-said thing, a physical object that carries tension, or a decision the character has to make. Describe each micro-conflict option in one sentence and explain how it would alter the passage's emotional trajectory. I will choose from the options you provide and revise accordingly.
Using AI to A/B Test a Sluggish Opening
One of the most underused AI workflows for novelists is comparative testing—running two versions of a chapter opening against each other to isolate which structural choices are responsible for the pacing difference before you commit to a full revision.
Here's how this works in practice. After your triage prompt identifies a problem in your chapter's opening pages, you write a brief alternate version yourself—not a polished draft, just a structural sketch where you've made different choices. Maybe you've entered the scene later, cutting the setup entirely. Maybe you've opened with dialogue instead of interiority. Maybe you've moved the reveal from the end of the opening to the beginning. It doesn't need to be good prose. It needs to be structurally different.
Then you run both versions through a comparison prompt that asks the AI to evaluate them against your stated chapter intent—not against each other aesthetically, but against the specific narrative job you need the chapter to do.
I have two versions of a chapter opening and I need help evaluating which one better serves the chapter's narrative purpose. I am not looking for a quality judgment. I need a functional assessment. Chapter intent: [paste your 1–3 sentence intent statement] Version A (original): [paste original opening—first 300–500 words] Version B (structural sketch): [paste alternate version—can be rough] Please evaluate both versions against the stated chapter intent using these specific criteria only: 1. Entry point: Which version enters the scene at the moment of highest potential energy for this chapter's purpose? Explain why. 2. Reader orientation: Which version more quickly tells the reader what is at stake in this scene without over-explaining it? 3. Momentum trajectory: Based on the opening alone, which version is more likely to maintain or increase the reader's forward pull through the rest of the chapter? What in the structure creates that trajectory? 4. Interiority balance: Which version uses the character's internal state more efficiently—present but not dominant? After the four-criteria breakdown, give me one sentence identifying which version you'd recommend pursuing for revision and the single most important structural reason why. Do not merge the versions. Do not rewrite either version. I need the comparison only.
What this prompt does that a simple "which is better" question can't is force the AI to evaluate structure against purpose. Version B might be rougher prose, but if it enters at the right moment and creates a momentum trajectory that Version A never achieves, that's what you need to know. The prose can be elevated in revision. A structural misfire can't be polished away.
A Note on What AI Can't Do Here
These workflows are genuinely useful, but they have a hard limit that's worth naming. AI can identify where momentum stalls and categorize the type of failure with reasonable accuracy. It cannot tell you whether a slow chapter should be fixed or whether it's doing quiet structural work your novel needs. That judgment belongs to you, and it requires knowing your book's architecture at a level no AI has access to.
The triage prompt will sometimes flag a deliberately quiet passage as low-momentum—because it is. That doesn't mean it's broken. It means you need to decide whether the stillness is earning its place. The diagnostic is a map, not a verdict. You're still the one who decides what the territory is supposed to look like.
Run these prompts on chapters you suspect are struggling, not on every chapter in the manuscript. The workflow is most valuable when your instinct is already telling you something's wrong and you need a framework for locating the specific failure rather than a general sense that the prose could be tighter. Used that way, it functions like a second reader who has no attachment to your pages—willing to say exactly where the forward motion stops, without flinching.
No comments yet. Be the first to comment!