Skills DirectorySkills Directory
SkillsLearnSecurityCategoriesDocsCommunityBlog
Sign InSubmit Skill
Skills Directory

Security-tested agent skills for Claude, coding agents, and AI workflows.

Directory

  • Browse Skills
  • All Skills A–Z
  • Claude Skills
  • Claude Code Skills
  • Agent Skills
  • Categories
  • Submit a Skill

Learn

  • Learn Hub
  • Install Claude Skills
  • Write SKILL.md
  • Skills vs MCP
  • Directories Compared

Security

  • Security
  • Methodology
  • Secure Claude Skills
  • Security Badges

Company

  • About
  • Community
  • Blog
  • API Docs
  • Advertise

2026 Skills Directory. All rights reserved.

Back to skills

Writing Without Slop

ASecurity

Source: ToxaBes, "The Unbearable Slop-ness of Being," Habr, https://habr.com/ru/articles/1063010/ Using a model is fine. Shipping the unedited draft is the failure. The reader tires of **recognizable patterns**, not of grammar. Frequency is the giveaway; a single instance is just ordinary writing. Applies to both Russian and English. The same rules apply to chat messages and files.

2 stars
0 votes
0 copies
0 views
Added 9/22/2026
ai-agentsapi

Works with

cliapi

Security Analysis

A100/100

Scanned 9/22/2026

Install to Claude Code

$npx -y skills add harnesys/studio --skill writing-without-slop --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Writing Without Slop?

Add the live security badge to your README — it updates automatically with every re-scan.

Security grade badge for Writing Without Slop
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/harnesys-writing-without-slop/badge)](https://www.skillsdirectory.com/skills/harnesys-writing-without-slop)

More formats (shields.io, HTML) on the badges page.

Download Zip
Files
SKILL.md
---
name: writing-without-slop
description: >
Use when writing or editing user-facing prose: chat replies, READMEs, VISION
  statements, DECISIONS, design docs, PR descriptions, explanatory comments, or
  any Markdown. Use when the text has an "AI accent," gritty readability,
  unedited model drafts, em-dash-heavy marketing cadences, "not-X-but-Y"
  structures, cliffhanger headings, filler phrases (e.g., "it is important to
  note," "let's dive in"), slogans, "rule-of-three" lists, or vague
  "research shows" claims. Use when the user runs `/writing-without-slop`.
---

# Writing without slop

Source: ToxaBes, "The Unbearable Slop-ness of Being," Habr, https://habr.com/ru/articles/1063010/

Using a model is fine. Shipping the unedited draft is the failure. The reader tires of **recognizable patterns**, not of grammar. Frequency is the giveaway; a single instance is just ordinary writing.

Applies to both Russian and English. The same rules apply to chat messages and files.

## Output contract

Every reply and every Markdown page follows this order:

1. The fact, decision, or action.
2. The supporting details the reader needs (names, numbers, identifiers, limits).
3. What you did not verify, if anything.

Use standard punctuation: periods, commas, colons, parentheses. Headings name the object (`ApiClient`, `handleRequest()`, `config.json`). A paragraph is 2–5 sentences long. A section ends with a fact, constraint, or example.

## Before / after

```
# before
The service isn't just an API; it's a powerful tool that changes the game.

Let's be honest: portability plays a key role.

It is important to note: this isn't about code, but about data. The system saves a snapshot—and that is the only way.
```

```
# after
The service compiles the JSON config into IR. The host saves a snapshot before `commit()`.

The config is portable as data. It does not store runtime state.
```

## Lexical markers (delete or replace)

RU: honest breakdown, let's be honest, no fluff, no holds barred, frankly speaking, spoiler, let's dive in, plays a key/decisive role, game-changer, important to note, worth emphasizing, unique, innovative, revolutionary, unsurpassed, powerful tool, undoubtedly, certainly, harmoniously, comprehensive approach, systematically, is [an assessment] instead of an action verb.

EN: delve, underscore, boast, meticulous, commendable, showcase, intricate, tapestry, let's dive in, deep dive, game-changer, plays a crucial/vital role, it's important to note, landscape, testament to, leverage, robust, seamless.

A human may use any of these once. A draft that stacks them is still a draft.

## Structural patterns

Edit for **frequency**, not a total ban. Generic examples use placeholders like `Module`, `handleRequest()`, `config.json`.

| # | Pattern | What to write instead |
|---|---|---|
| 1 | Em dash (`—` / `--`) as the default comma, colon, or parenthesis; space before ` —` | Keep at most one real break of thought per several paragraphs. Elsewhere: comma, colon, parentheses, or two sentences. |
| 2 | Antithesis `not X, but Y` / `it's not X, it's Y` / `it's not about X` | State Y. At most one contrast per document, at the strongest point; never as the first sentence of a section. |
| 3 | One-line paragraph for fake drama (`And then everything went wrong.`) | One-sentence paragraphs only at a genuine turning point. Otherwise, merge with the adjacent paragraph. |
| 4 | Staccato negations: series of single-sentence `No X.` / `It's not Y.` / `It does nothing.` | Merge into 2–3 sentence paragraphs including the reason or limitation. A single bare negation can close a section; five of them create a slogan. |
| 5 | Cliffhanger headings ("A crack in the foundation," "The hidden cost," "Is it worth using?") | Heading = noun or artifact the section is about (`ApiClient`, `handleRequest()`, "When to use the module"). Check headings as a list separate from the body. Avoid question headings for reference docs. |
| 6 | Filler ("essentially," "roughly speaking," "seemingly," "thus," "ultimately," "furthermore") | Delete if the sentence retains the same meaning. Keep only a genuine caveat or a real transition. |
| 7 | Aphoristic closer after every block | Close with the next fact, a limit, a question answered by the next section, or an identifier (`status: done`, `limit: 100`, `file: config.json`). |
| 8 | Negation closer: every section ends with "X is not included," "Y does not do," "Z does not exist" | End half the sections with what the component *does*, its inputs/outputs, or its limit. A bare "not included" list without grouping is a telltale sign. |
| 9 | Forced symmetry and rule of three ("innovative, transformative, groundbreaking") | List as many items as actually exist (2, 4, 5). Break parallel syntax once: one short item, one long. |
| 10 | Concreteness decay: names and numbers in the first third, slogans in the last | Reread the last third in isolation. Ensure the same density of IDs, numbers, and examples there. A long exclusion list without IDs or commands represents decay. |
| 11 | Vague authority ("studies show," "experts note," "marked a turning point") | Name the source, file, identifier, or measurement. If you cannot, cut the sentence. Empty authority.

Attribution

harnesysharnesys
View sourceMore from harnesys →
SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments (0)

No comments yet. Be the first to comment!

SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Related Skills

Caveman

Ultra-compressed communication mode that cuts output tokens while keeping technical accuracy. Levels: lite, full, ultra and the wenyan variants. Use for /caveman, "caveman mode", "talk like caveman", "be brief" or "less tokens".

1066601 votes

Hyperplan

Adversarial multi-agent planning skill. Self-orchestrates 5 hostile category members (unspecified-low, unspecified-high, deep, ultrabrain, artistry) via team-mode for ruthless cross-critique debate, distills only the defensible insights, then MANDATORILY hands the distilled insight bundle to the `plan` agent for executable plan formalization. Use when planning needs maximum rigor and surfacing of weak assumptions, blind spots, and over-engineering. Triggers: 'hyperplan', 'hpp', '/hyperplan', ...

686011 votes

Mcp Code Execution

Routes multi-tool workflows through MCP servers for large datasets and pipelines. Use when Bash tool overhead is limiting throughput on data-heavy tasks.

3351 votes

catchup

Recovers the conversation and failed tool calls of a previous Codex, Claude Code, Antigravity, Cline, Copilot CLI, Cursor, DeepSeek Harness, Kimi, OpenCode, Pi Agent, or ZCode session. Use when the user says "catch up", "what did the last session do", "get me up to speed", "I switched agents", asks to recover/summarize a previous session before continuing, or asks to diagnose or report a catchup failure. Do NOT use for the current conversation, git history, or any non-agent log.

651 votes

math-skill

A comprehensive mathematical reasoning skill for AI assistants — handles arithmetic to research-level problems with rigorous step-by-step reasoning, systematic verification, and transparent uncertainty handling

381 votes
View all in ai-agents →