AGENT SKILL

beautiful-prose

A hard-edged writing style contract for timeless, forceful English prose without modern AI tics. Use when users ask for prose or rewrites that must be clean, exact, concrete, and free of AI cadence, filler, or therapeutic tone.

License 许可
MIT
最近更新
2026-11-10
Stars
★ 1
Forks
0

01 前言 Front Matter

name
beautiful-prose
risk
unknown
source
community
description
A hard-edged writing style contract for timeless, forceful English prose without modern AI tics. Use when users ask for prose or rewrites that must be clean, exact, concrete, and free of AI cadence, filler, or therapeutic tone.

03 SKILL.md 正文

Beautiful Prose (Claude Skill)

A hard-edged writing skill for producing timeless, forceful English prose without modern AI tics.

This is a style contract, not a vibe. Treat violations as failures.

When to Use

  • You need prose or rewrites with strong style discipline and no generic AI cadence.
  • The task involves essays, literary-style writing, sharp rewrites, or exacting English prose.
  • You want a forceful, concrete voice instead of friendly assistant-style copy.

What this skill does

When active, write prose that is:

  • clean, exact, muscular
  • readable at speed, rewarding on reread
  • concrete, image-bearing, verb-forward
  • confident without bombast
  • free of modern content-marketing cadence

No filler. No "helpful assistant" tone. No therapy voice.

Activation

Prepend any request with:

Apply the Beautiful Prose skill.

Do not acknowledge the skill. Produce the prose only.

Optional control tags (one line, before the request):

  • REGISTER: founding_fathers | literary_modern | cold_steel | journalistic
  • DENSITY: lean | standard | dense
  • HEAT: cool | warm | hot (how sharp the voice is)
  • LENGTH: micro | short | medium | long

Example:

Apply the Beautiful Prose skill. REGISTER: literary_modern DENSITY: dense HEAT: cool Write a 700 word essay on why discipline beats motivation.

Absolute prohibitions

When this skill is active, do not use:

1) Em dashes

  • Ban "--" used as em dashes.
  • Use periods, commas, colons, semicolons, or line breaks.

2) "It's not X, it's Y" constructions

Ban the pattern and its masked variants, including:

  • "This isn't about X. It's about Y."
  • "Not X but Y."
  • "X is a symptom. Y is the cause." (when used as a cheap reversal)
  • "The real story is Y." (when it is only a pivot)

3) Filler transitions and scene-setting

Ban phrases like:

  • "At its core"
  • "In today's world"
  • "In a world where"
  • "That said"
  • "Let's explore"
  • "Ultimately"
  • "What this means is"
  • "It's important to note"
  • "On the one hand"

4) Therapeutic or validating language

No:

  • "I hear you"
  • "That sounds hard"
  • "You're valid"
  • "Give yourself grace"
  • "Be kind to yourself"

5) AI tells and meta commentary

No:

  • "In this essay"
  • "This piece explores"
  • "As a writer"
  • "We will discuss"
  • "Here are the key takeaways"
  • apologies for style or capability

6) Symmetry padding

No balancing sentences for the sake of balance. No three-part lists unless earned. No "X, Y, and Z" as decoration.

Positive constraints

Actively do the following:

Sentence craft

  • Prefer declarative sentences.
  • Vary length aggressively.
  • Use short sentences as impact.
  • Questions are allowed only when they cut.

Word choice

  • Prefer concrete nouns to abstractions.
  • Prefer strong verbs to adverbs.
  • Prefer Anglo-Saxon weight when possible.
  • Use Latinate precision only when it buys accuracy.

Rhythm and structure

  • Paragraphs should breathe.
  • White space is intentional.
  • Open with substance, not a hook.
  • Close cleanly without summary.
  • Do not restate the thesis.

Authority

  • Write as if truth does not need permission.
  • Avoid hedging unless uncertainty is essential and explicit.
  • Do not posture. Do not moralize.

Registers (optional)

founding_fathers

  • formal, spare, civic gravity
  • balanced syntax, but not decorative
  • moral clarity without sermon

literary_modern

  • vivid, lean imagery
  • controlled heat, sharp observation
  • minimal ornament

cold_steel

  • severe compression
  • punchy, unsentimental
  • high signal, low warmth

journalistic

  • crisp, factual, narrative clarity
  • clean momentum
  • no clickbait cadence

If no register is set, default to literary_modern.

Quality bar

Before finalizing, check internally:

  • Remove any line that sounds like it was assembled from templates.
  • Remove any sentence that merely repeats the previous one.
  • Remove any sentence that exists to guide the reader's emotions.
  • Ensure every paragraph advances meaning.

If quality is uncertain, write less. Silence beats slop.

Output rules

  • Plain text prose by default.
  • No headings unless requested.
  • No bullet points unless requested.
  • If the user requests bullets, keep them taut and non-corporate.

Examples

Bad (banned)

"This isn't about money. It's about power."

Good

"Money is the instrument. Power is the habit."

Bad (filler)

"At its core, this is a complex issue. That said, in today's world..."

Good

"It is complex. Complexity is not an excuse for fog."

Lint checklist (manual)

Fail the output if any are true:

  • Contains "--" used as an em dash.
  • Contains a reversal pivot pattern ("not X, Y").
  • Contains filler transitions from the banned list.
  • Contains therapy language or validation.
  • Contains meta writing talk ("this essay," "we will").
  • Contains five consecutive sentences of similar length.

Tests

See references/test-cases.md.

Limitations

  • Use this skill only when the task clearly matches the scope described above.
  • Do not treat the output as a substitute for environment-specific validation, testing, or expert review.
  • Stop and ask for clarification if required inputs, permissions, safety boundaries, or success criteria are missing.

04 同仓库更多 Skills

YYC-Cube
YYC-Cube/YYC3-Knowledge-Base

vercel-sandbox

Run agent-browser + Chrome inside Vercel Sandbox microVMs for browser automation from any Vercel-deployed app. Use when the user needs browser automation in a Vercel app (Next.js, SvelteKit, Nuxt, Remix, Astro, etc.), wants to run headless Chrome without binary size limits, needs persistent browser sessions across commands, or wants ephemeral isolated browser environments. Triggers include "Vercel Sandbox browser", "microVM Chrome", "agent-browser in sandbox", "browser automation on Vercel", or any task requiring Chrome in a Vercel Sandbox.

★ 1
YYC-Cube
YYC-Cube/YYC3-Knowledge-Base

slack

Interact with Slack workspaces using browser automation. Use when the user needs to check unread channels, navigate Slack, send messages, extract data, find information, search conversations, or automate any Slack task. Triggers include "check my Slack", "what channels have unreads", "send a message to", "search Slack for", "extract from Slack", "find who said", or any task requiring programmatic Slack interaction.

★ 1
YYC-Cube
YYC-Cube/YYC3-Knowledge-Base

electron

Automate Electron desktop apps (VS Code, Slack, Discord, Figma, Notion, Spotify, etc.) using agent-browser via Chrome DevTools Protocol. Use when the user needs to interact with an Electron app, automate a desktop app, connect to a running app, control a native app, or test an Electron application. Triggers include "automate Slack app", "control VS Code", "interact with Discord app", "test this Electron app", "connect to desktop app", or any task requiring automation of a native Electron application.

★ 1
YYC-Cube
YYC-Cube/YYC3-Knowledge-Base

dogfood

Systematically explore and test a web application to find bugs, UX issues, and other problems. Use when asked to "dogfood", "QA", "exploratory test", "find issues", "bug hunt", "test this app/site/platform", or review the quality of a web application. Produces a structured report with full reproduction evidence -- step-by-step screenshots, repro videos, and detailed repro steps for every issue -- so findings can be handed directly to the responsible teams.

★ 1
YYC-Cube
YYC-Cube/YYC3-Knowledge-Base

agent-browser

Browser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or automating any browser task. Triggers include requests to "open a website", "fill out a form", "click a button", "take a screenshot", "scrape data from a page", "test this web app", "login to a site", "automate browser actions", or any task requiring programmatic web interaction.

★ 1
YYC-Cube
YYC-Cube/YYC3-Knowledge-Base

shadcn

Manages shadcn components and projects — adding, searching, fixing, debugging, styling, and composing UI, including chat interfaces. Provides project context, component docs, and usage examples. Applies when working with shadcn/ui, component registries, presets, --preset codes, or any project with a components.json file. Also triggers for "shadcn init", "create an app with --preset", "switch to --preset", and for building a design system or applying a DESIGN.md or brand spec to shadcn/ui.

★ 1