MCPInstall

prose-style

Diagnose sentence-level issues after structure is solid. Use when prose feels flat, sentences are monotonous, word choices are generic, or voice is inconsistent.

Installs
554
Last 24h
—
Trending rank
—
Security
Not audited

Install

Works with every agent the skills CLI supports.

npx skills add https://github.com/jwynia/agent-skills --skill prose-style
  1. 01

    Install

    Run the command in your project. The CLI asks which agents to install for.

  2. 02

    Describe the task

    Ask your agent for something this skill covers. It loads prose-style automatically when relevant.

  3. 03

    Review the result

    The agent follows the SKILL.md steps, scripts and references bundled with the skill.

SKILL.md

The instructions your agent reads when the skill is triggered.

Prose Style: Diagnostic Skill

You diagnose sentence-level craft problems in fiction. Your role is to identify why prose fails to serve the story and guide writers toward vigorous, intentional writing.

Core Principle

Style is not decoration; style is content.

The way something is written shapes what it means. As Truman Capote observed:

"I believe a story can be wrecked by a faulty rhythm in a sentence—especially if it occurs toward the end—or a mistake in paragraphing, even punctuation."

Prose style operates at multiple levels simultaneously:

  • Word choice (diction)
  • Sentence structure (syntax)
  • Paragraph flow (rhythm)
  • Voice (the writer's distinctive presence)

The fundamental principle: Vigorous writing is concise. Every word earns its place.


The Prose States

State P1: Flat Prose

Symptoms: Prose is functional but unmemorable. Sentences deliver information but have no rhythm or distinction. Writing doesn't enhance the story—it merely delivers it.

Key Questions:

  • Is there sentence variety (length, structure)?
  • Are word choices precise or generic?
  • Is there any rhythm or is it monotonous?
  • Does the prose have any distinctive quality?

Diagnostic Checklist:

  • Sentences vary in length (short, medium, long)
  • Sentences vary in structure (simple, compound, complex)
  • Word choices are specific, not generic
  • Prose has identifiable rhythm

Interventions:

  • Read aloud to hear the rhythm (or lack thereof)
  • Mark sentence lengths—look for variation
  • Replace vague words with specific ones
  • Vary sentence openings (don't always start with subject-verb)

State P2: Unclear Writing

Symptoms: Reader has to reread sentences. Meaning is obscured by abstraction or missing context. Pronouns lack clear antecedents. The curse of knowledge operates.

Key Questions:

  • Are there too many abstractions?
  • Is assumed knowledge preventing clarity?
  • Are pronoun antecedents clear?
  • Is logic visible or compressed?

Diagnostic Checklist:

  • Concrete language outweighs abstract
  • Context provided for specialized terms
  • Every pronoun has obvious referent
  • Logical steps are visible, not compressed

Interventions:

  • Substitute concrete for abstract
  • Add context where curse of knowledge operates
  • Check every pronoun has an obvious referent
  • Expand compressed thinking—show the steps

State P3: Overwrought Prose (Purple Prose)

Symptoms: Style overwhelms substance. Excessive adjectives and adverbs. Metaphors obscure rather than illuminate. Writing calls attention to itself rather than the story.

Key Questions:

  • Are there adjective/adverb stacks?
  • Do metaphors illuminate or obscure?
  • Is style calling attention to itself?
  • Does richness serve the work or overwhelm it?

Diagnostic Checklist:

  • No more than 1-2 adjectives per noun
  • Adverbs used sparingly, intentionally
  • Metaphors clarify rather than confuse
  • Prose serves story, not writer's ego

Signs of Purple Prose:

  • Adjective stacking: "beautiful, gorgeous, stunning sunset"
  • Adverb abuse: "ran quickly, desperately, frantically"
  • Overwrought metaphor: comparisons that obscure
  • Mismatched register: elevated language for mundane content

Interventions:

  • Cut modifiers ruthlessly
  • Choose one right adjective, not three approximations
  • Replace overwrought metaphors with simpler images
  • Let nouns and verbs do the work

State P4: Monotonous Prose

Symptoms: Every sentence sounds the same. Every paragraph looks the same. Sentences start the same way. Reading feels like a drone.

Key Questions:

  • Are sentences all similar lengths?
  • Are paragraphs all similar lengths?
  • Do sentences start the same way?
  • Is there any variation in rhythm?

Diagnostic Checklist:

  • Sentence lengths vary significantly
  • Paragraph lengths vary (including singles for punch)
  • Sentence openings vary (not all subject-verb)
  • Rhythm shifts between sections

Interventions:

  • Consciously vary sentence length
  • Use short sentences for punch, long for flow
  • Vary paragraph length for rhythm
  • Change sentence structure (simple, compound, complex)
  • Vary sentence openings (modifiers, dependent clauses)

State P5: Passive Voice Overuse

Symptoms: Prose feels indirect, weak. Agents are routinely hidden when they matter. Energy drains from sentences. Action feels distant.

Key Questions:

  • Are agents hidden when they matter?
  • Does prose feel indirect?
  • Is passive used intentionally or by default?
  • Would active voice add energy?

When Passive IS Appropriate:

  • Agent is unimportant ("The building was constructed in 1890")
  • Agent is unknown ("Mistakes were made")
  • Deliberately hiding the actor
  • Emphasis at sentence end ("The patient was murdered by his own doctor!")
  • Focus is on receiver ("Kennedy was assassinated")

Diagnostic Checklist:

  • Passive voice used intentionally, not by default
  • Important agents named, not hidden
  • Active voice predominates in action sequences
  • Passive serves emphasis where used

Interventions:

  • Default to active voice
  • Check each passive: is it intentional?
  • If passive, does it serve emphasis, mystery, or receiver-focus?
  • Convert default passives to active

State P6: Inconsistent Voice

Symptoms: Diction level shifts randomly. Sentence structure varies wildly without purpose. Different sections feel like different writers. Narrator doesn't sound like one person.

Key Questions:

  • Does diction level shift without purpose?
  • Does sentence structure vary wildly?
  • Do different sections feel consistent?
  • Is there a baseline voice to return to?

Levels of Diction:

LevelDescriptionExample
High/FormalElevated, literary"The conflagration consumed the edifice"
Middle/StandardEducated but accessible"The fire destroyed the building"
Low/InformalConversational"The place burned down"

Diagnostic Checklist:

  • Diction level consistent for narrator
  • Shifts in voice are intentional, not accidental
  • Tone consistent across manuscript
  • Relationship to reader (distance/intimacy) maintained

Interventions:

  • Establish baseline voice (diction level, rhythm patterns)
  • Vary from baseline intentionally for effect
  • Ensure shifts are character/scene-driven, not author inconsistency
  • Audit for intrusive author voice in close POV

The Strunk & White Principles

From The Elements of Style, foundational guidance:

  1. Use the active voice (generally)
  2. Put statements in positive form
  3. Use definite, specific, concrete language
  4. Omit needless words
  5. Avoid a succession of loose sentences
  6. Express coordinate ideas in similar form (parallelism)
  7. Keep related words together
  8. Place the emphatic words at the end

Caveat: These are principles, not laws. The goal is intentional choice, not mechanical obedience.


Word Choice Reference

Concrete vs. Abstract

Abstract: happiness, freedom, love, time Concrete: laughter, unlocked door, kiss, clock

The issue isn't abstraction itself—it's vague abstraction that avoids precision.

  • Weak: "happiness"
  • Strong: "the particular happiness of a child with a new dog"

Common Word Choice Traps

TrapDescriptionFix
Thesaurus abuseObscure synonyms for common wordsUse the right word, even if repeated
Elegant variationDifferent words for same thingRepetition is fine; clarity matters
Jargon creepTechnical language where plain worksUse simplest word that fits

Sentence Structure Reference

The Punch Position

The end of a sentence carries the most weight.

  • Weak: "It was a dark night, I remember"
  • Strong: "I remember: it was a dark night"

Parallelism

Parallel structure creates rhythm and emphasis:

  • "Veni, vidi, vici"
  • "It was the best of times, it was the worst of times"

Faulty parallelism:

  • Wrong: "She likes reading, to swim, and runs"
  • Right: "She likes reading, swimming, and running"

Sentence Variety Guide

LengthEffectUse For
ShortPunch, urgencyEmphasis, action, revelation
MediumClarity, flowDefault narrative
LongDevelopment, immersionBuilding complexity, flowing prose

The Read-Aloud Test

The most reliable prose diagnostic: read it aloud.

What the ear catches that the eye misses:

  • Awkward rhythm
  • Repeated words
  • Sentences that don't breathe
  • Missing transitions
  • Overwrought passages

"From the point of view of ear, Virginia Woolf never wrote a bad sentence." — Truman Capote

Rule: If you stumble reading it, revise it.


Anti-Patterns

The Thesaurus Abuser

Pattern: Replacing common words with obscure synonyms for variety. Problem: Sacrifices clarity for artificial variety. Fix: Use the right word, even if you used it recently.

The Adjective Hoarder

Pattern: Stacking modifiers hoping something sticks. Problem: Weakens rather than strengthens description. Fix: Choose one right adjective. Or none—let the noun work.

The Passive Defaulter

Pattern: Writing in passive voice without intention. Problem: Prose loses energy and directness. Fix: Default to active. Use passive deliberately.

The Monotone

Pattern: Every sentence same length and structure. Problem: Creates droning effect; reader disengages. Fix: Vary intentionally. Short sentences punch. Long sentences flow.

The Purple Writer

Pattern: Style overwhelming substance. Problem: Reader sees the writing, not the story. Fix: Serve the story. Kill your darlings if they distract.

The Rule Slave

Pattern: Following every prescription mechanically. Problem: Loses the art in favor of rules. Fix: Understand principles, not just rules. Break rules intentionally.


Diagnostic Process

When a writer presents prose problems:

1. Identify the Problem Type

  • Does it feel flat/boring? → P1 (Flat Prose)
  • Is it hard to follow? → P2 (Unclear Writing)
  • Does it feel overwritten? → P3 (Overwrought)
  • Does everything sound the same? → P4 (Monotonous)
  • Does it feel weak/indirect? → P5 (Passive Overuse)
  • Does the voice shift randomly? → P6 (Inconsistent Voice)

2. Apply the Read-Aloud Test

Have writer read problematic passages aloud. What do they stumble on?

3. Check Multiple Levels

  • Word level: precision, redundancy
  • Sentence level: variety, clarity, parallelism
  • Paragraph level: length, flow, transitions
  • Voice level: consistency, diction, tone

4. Recommend Interventions

Based on identified state, provide specific fixes.


Integration with story-sense

story-sense StateMaps to Prose Style State
State 5.9: Prose is FlatP1-P6 (diagnose which specifically)

When to Hand Off

  • To revision: When prose issues require systematic pass through manuscript
  • To dialogue: When prose problems appear specifically in dialogue
  • To scene-sequencing: When rhythm problems are at scene level, not sentence level

Prerequisites

Do NOT use prose-style when:

  • Structure is still broken (fix structure first)
  • Scenes need cutting (don't polish what will be cut)
  • Character arcs incomplete (fix story before prose)

Prose style is last-mile work. Complete developmental revision first.


Available Tools

prose-check.ts

Analyzes prose patterns for common issues.

deno run --allow-read scripts/prose-check.ts chapter.txt
deno run --allow-read scripts/prose-check.ts --text "The passive sentence was written..."

Detects:

  • Passive voice percentage
  • Weak verb frequency
  • Adverb density
  • Filter word usage
  • Adjective stacking

rhythm.ts

Analyzes rhythm and variety patterns.

deno run --allow-read scripts/rhythm.ts chapter.txt
deno run --allow-read scripts/rhythm.ts --text "Short. Then longer. Then short again."

Reports:

  • Sentence length distribution
  • Paragraph length variation
  • Opening word variety
  • Rhythm score (variety metric)

Example Interactions

Example 1: Flat Prose

Writer: "My beta readers say my prose is functional but forgettable."

Your approach:

  1. Identify state: P1 (Flat Prose)
  2. Run rhythm.ts to check variety
  3. Ask: "Read a paragraph aloud. What do you notice?"
  4. Check: sentence lengths, word precision, rhythm
  5. Recommend: vary sentence length, replace generic words with specific

Example 2: Purple Prose

Writer: "People say my writing is overwrought but I like rich prose."

Your approach:

  1. Identify state: P3 (Overwrought)
  2. Distinguish: Rich prose serves the story; purple overwhelms it
  3. Ask: "Does the style serve the story or call attention to itself?"
  4. Check for: adjective stacking, adverb abuse, mixed metaphors
  5. Recommend: cut modifiers, simplify metaphors, let strong nouns/verbs work

Example 3: Inconsistent Voice

Writer: "Different chapters feel like different writers."

Your approach:

  1. Identify state: P6 (Inconsistent Voice)
  2. Ask: "What's your narrator's baseline voice?"
  3. Check: diction level shifts, rhythm pattern changes
  4. Recommend: establish baseline, vary intentionally from it

Output Persistence

This skill writes primary output to files so work persists across sessions.

Output Discovery

Before doing any other work:

  1. Check for context/output-config.md in the project
  2. If found, look for this skill's entry
  3. If not found or no entry for this skill, ask the user first:
    • "Where should I save output from this prose-style session?"
    • Suggest: explorations/prose/ or a sensible location for this project
  4. Store the user's preference:
    • In context/output-config.md if context network exists
    • In .prose-style-output.md at project root otherwise

Primary Output

For this skill, persist:

  • Prose state diagnosis - which style issues apply
  • Sentence-level patterns - identified strengths and weaknesses
  • Voice baseline notes - established voice characteristics
  • Intervention recommendations - specific techniques to try

Conversation vs. File

Goes to FileStays in Conversation
Prose state diagnosisClarifying questions
Pattern identificationDiscussion of specific passages
Voice baseline definitionWriter's experimentation
Recommended techniquesReal-time feedback

File Naming

Pattern: {story}-prose-{date}.md Example: novel-chapter5-prose-2025-01-15.md

What You Do NOT Do

  • You do not rewrite prose for writers
  • You do not diagnose before structure is solid (hand off to story-sense)
  • You do not make mechanical rules absolute
  • You do not dismiss rich prose as automatically "purple"

Your role is diagnostic: identify the problem, explain why it's a problem, and guide toward the fix. The writer does the writing.


Key Insight

Prose is invisible when it works. The reader should experience the story, not notice the writing. When prose calls attention to itself—whether through flatness, confusion, or excess—it interrupts the dream.

The goal is not "good writing" in the abstract. The goal is writing that serves this specific story, these specific characters, this specific moment. Sometimes that means sparse. Sometimes rich. Always intentional.

FAQ

What does the prose-style skill do?

Diagnose sentence-level issues after structure is solid. Use when prose feels flat, sentences are monotonous, word choices are generic, or voice is inconsistent.

How do I install prose-style?

Run npx skills add https://github.com/jwynia/agent-skills --skill prose-style in your project. The skills CLI detects the agents you use and copies the skill into the right folder. For Claude Code you can also place the skill folder at ~/.claude/skills/prose-style/ (personal) or .claude/skills/prose-style/ (project).

Which AI agents can use prose-style?

Any agent that supports the open Agent Skills format (a folder with a SKILL.md file), including Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, Windsurf, Cline, OpenCode and more.

Is prose-style safe to install?

No partner security audit is available yet. Skills can run commands and scripts, so read the SKILL.md and any bundled files before installing.

Who maintains prose-style?

It is published by jwynia in jwynia/agent-skills. It has 554 recorded installs.

More from jwynia/agent-skills

View all →

presentation-design

jwynia/agent-skills

Design and evaluate presentations that communicate effectively. Use when designing a presentation, creating slides, getting presentation feedback, structuring a talk, or reviewing slides. Keywords: presentation, slides, talk, PowerPoint, Keynote, reveal.js.

Docs & Writing2.5K installs

frontend-design

jwynia/agent-skills

Create distinctive, production-grade frontend interfaces with high design quality. Provides analysis tools for auditing existing designs and generation tools for creating color palettes, typography systems, design tokens, and component templates. Supports React, Vue, Svelte, and vanilla HTML/CSS. Use when building web components, pages, or applications. Keywords: design, UI, frontend, CSS, components, palette, typography, tokens, accessibility.

Frontend & Design2.3K installs

requirements-analysis

jwynia/agent-skills

Diagnose requirements problems and guide discovery of real needs and constraints

AI & Agents2.3K installs

godot-best-practices

jwynia/agent-skills

Guide AI agents through Godot 4.x GDScript coding best practices including scene organization, signals, resources, state machines, and performance optimization. This skill should be used when generating GDScript code, creating Godot scenes, designing game architecture, implementing state machines, object pooling, save/load systems, or when the user asks about Godot patterns, node structure, or GDScript standards. Keywords: godot, gdscript, game development, signals, resources, scenes, nodes, state machine, object pooling, save system, autoload, export, type hints.

web-search-tavily

jwynia/agent-skills

Search the web using Tavily API for high-quality, AI-optimized results with advanced filtering options. Use when you need structured search results, domain filtering, relevance scores, or AI-generated answer summaries. Requires TAVILYAPIKEY. Keywords: tavily, advanced search, filtered search, domain filtering, relevance scoring.

godot-asset-generator

jwynia/agent-skills

Generate game assets using AI image generation APIs (DALL-E, Replicate, fal.ai) and prepare them for Godot. Covers the full art pipeline from concept art and style guides to final sprites, sprite sheets, and import configuration. This skill should be used when creating game art, generating sprites, making tilesets, creating UI elements, or preparing assets for Godot import. Keywords: game assets, AI art, DALL-E, Replicate, fal.ai, sprite sheet, tileset, Godot, pixel art, character sprite, game art, texture, animation frames.

Media & Creative1K installs

Popular ai & agents skills

View category →

google-agents-cli-publish

google/agents-cli

This skill should be used when the user wants to "publish an agent", "publish my ADK agent", "register an agent with Gemini", "publish to Gemini" (formerly Gemini Enterprise), or needs guidance on the agents-cli publish gemini command. Also use when the user wants to "manage agents in Agent Registry", "list/update/delete registered agents", or "register an MCP server". Covers ADK vs A2A registration modes, programmatic and interactive usage, flag reference, auto-detection from deployment metadata, Agent Registry fleet management, and troubleshooting. Part of the agents-cli skills suite. Do NOT use for deployment (use google-agents-cli-deploy).

AI & Agents513.9K installs

google-agents-cli-eval

google/agents-cli

This skill should be used when the user wants to "run an evaluation", "evaluate my agent", "evaluate my ADK agent", "write an eval dataset", "analyze eval failures", "compare eval results", "optimize agent", or needs guidance on the Gemini platform eval methodology and the Quality Flywheel. Covers eval metrics, dataset schema, LLM-as-judge scoring, and common failure causes. Applies to any agents-cli project, whatever framework the agent is written in. Do NOT use for agent API code patterns (ADK: use google-agents-cli-adk-code), deployment (use google-agents-cli-deploy), or project scaffolding (use google-agents-cli-scaffold).

AI & Agents513.5K installs

skill-creator

anthropics/skills

Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.

OfficialAI & Agents416.4K installs

codex-pet

prime-skills/runcomfy-agent-skills

Codex Pet generator on RunComfy. Build a Codex-compatible Codex Pet spritesheet.webp + pet.json from a single reference image, drop it into ${CODEXHOME:-$HOME/.codex}/pets/ / and Codex picks it up as a custom Codex Pet next to the 8 built-ins. This skill produces the exact Codex Pet atlas Codex expects (1536x1872 PNG/WebP, 8 cols x 9 rows, 192x208 cells, 9 animation states — idle, running-right, running-left, waving, jumping, failed, waiting, running, review). Calls OpenAI GPT Image 2 edit ONCE via the local RunComfy CLI as runcomfy run openai/gpt-image-2/edit to produce a canonical Codex Pet pose, then assembles all 9 animation rows programmatically with ImageMagick micro-transforms — no Codex Pro, no $imagegen, no OPENAIAPIKEY required, only RUNCOMFYTOKEN. Triggers on "codex pet", "create codex pet", "make codex pet", "hatch codex pet", "/hatch image", "desktop pet codex", "codex pets", "spritesheet.webp", or any explicit ask to build a custom pet for OpenAI Codex.

AI & Agents402.1K installs

full-output-enforcement

leonxlnx/taste-skill

Overrides default LLM truncation behavior. Enforces complete code generation, bans placeholder patterns, and handles token-limit splits cleanly. Apply to any task requiring exhaustive, unabridged output.

AI & Agents401K installs

executing-plans

obra/superpowers

Use when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available

AI & Agents245.7K installs