Skip to content
Skill

writing-style

by DheerG

AI Summary

A sentence or paragraph can look fine in isolation. When the same structural shape repeats across 1,000-3,000 words, the aggregate creates a mechanical feeling — even though no single instance is wrong. Linear readers miss this because they process sequentially and each paragraph clears the "is this

Install

Copy this and paste it into Claude Code, Cursor, or any AI assistant:

I want to install the "writing-style" skill in my project.

Please run this command in my terminal:
# Install skill into your project
mkdir -p .claude/skills/writing-style && curl --retry 3 --retry-delay 2 --retry-all-errors -o .claude/skills/writing-style/SKILL.md "https://raw.githubusercontent.com/DheerG/swarms/main/skills/writing-style/SKILL.md"

Then restart Claude Code (or reload the window in Cursor) so the skill is picked up.

Description

Structural pattern analysis for writing-mode review. Performs paragraph-level decomposition and pattern aggregation. Invoked by the team lead during editorial review.

Purpose

A sentence or paragraph can look fine in isolation. When the same structural shape repeats across 1,000-3,000 words, the aggregate creates a mechanical feeling — even though no single instance is wrong. Linear readers miss this because they process sequentially and each paragraph clears the "is this well-written?" bar individually. This analysis solves the forest-vs-trees problem. It decomposes every paragraph's structure, then aggregates the fingerprints to surface repetition invisible to linear reading.

How to Use

Invoke this skill using the Skill tool: swarm:writing-style. Pass the piece's text or file path as the args parameter. The skill returns a structured report. The lead relays the report to the editor for interpretation. Invoke during the review phase (required) or optionally during drafting. The report is advisory, never gating. The editor applies editorial judgment to the data. The Economist escape clause applies: if a flagged pattern serves the piece, keep it and note why.

Analysis Process

Exclude from analysis: Table cells, comparison tables, and structured list items with deliberately parallel construction (numbered checklists, step-by-step procedures). These are exempt from construction density calculation but still included in named pattern counts. Report construction density both with and without these items so the editor has both numbers. Introductory and concluding paragraphs adjacent to lists are NOT exempt — only the parallel list items themselves.

Pass 1: Structural Decomposition

For each body paragraph, classify: • Opening move — how the paragraph begins: • claim: declarative assertion • scenario: narrative with actors or concrete situation • question: interrogative opening • data: leads with a statistic or cited fact • contrast: opens with negation or opposition • definition: introduces or defines a term • transition: bridges from previous section • imperative: directive opening • Dominant sentence shape: • declarative: simple subject-verb-object statements • contrast-pivot: sentences built on "not X — it is Y" or "X, but Y" • question-answer: rhetorical question followed by its answer • conditional: "when X, then Y" or "if X, then Y" • enumeration: listing or cataloguing items • narrative: sequential events with actors • compound-aside: sentences with em-dash parenthetical asides (banned — flag for rewrite) • Named trope patterns present (if any): • negative-parallelism: "X is not [assumption]. It is [reframe]." • countdown: "Not X. Not Y. [Resolution]." • self-posed-qa: Question posed and immediately answered. Exclude H2/H3 section headings that are questions — question-format headings are a structural choice, not a rhetorical trope. Only count questions posed within body paragraph text. • same-phrase-recycled: A multi-word formulation appearing verbatim elsewhere in the piece • punchy-fragment: One-sentence standalone paragraph used for emphasis • dead-metaphor: A metaphor that has already appeared earlier in the piece • anaphora: Same opening word as the previous paragraph • tricolon: Three-part list or phrase Record each paragraph's structural fingerprint in compact notation.

Discussion

0/2000
Loading comments...

Health Signals

MaintenanceCommitted Today
Active
AdoptionUnder 100 stars
77 ★ · Niche
DocsREADME + description
Well-documented

GitHub Signals

Stars77
Forks6
Issues0
UpdatedToday
View on GitHub
MIT License

My Fox Den

Community Rating

Sign in to rate this booster

Works With

Claude Code