Unslop Skill · Writing and analysis

Unslop Skill: Build a Domain-Specific Writing Profile

Give a domain a profile your agent can carry forward.

Matt Shumer's Unslop Skill wraps a Python CLI that takes a domain, generates a prompt set and text or visual samples with Claude Code, analyzes the set, and writes analysis.md plus a reusable skill.md profile. It is a profile generator, separate from Stop Slop and Humanizer.

View Matt Shumer’s source

Domain profiles
Style analysis
Reusable skill generation

About This Skill

Unslop is Matt Shumer's domain-profile generator. Give its Python CLI a domain such as `blog writing`; it generates a prompt set and text or visual samples, then produces analysis.md and a reusable skill.md profile. It does not directly rewrite a draft and should not be merged with Stop Slop or Humanizer.

Built on proven foundations

Trusted Source, Professionally Refined

The original expertise stays intact. Focused refinements make it easier to start, follow the guidance, and know when the work is finished.

Original foundation

Matt Shumer · commit edcb623

The original Skill knowledge, package structure, attribution and original license terms are preserved. Follow the source link to see where this edition began.

View the original project

Antigravity optimization

6 files, one clearer workflow

A source-faithful edition with the improvements explained below. The download includes the original attribution and everything needed to follow the Skill’s guidance.

More useful from the first prompt

What We Improved

Keep the depth you came for. Get clearer starting points, practical guidance and an explicit finish line.

A domain-profile entity boundary

Use a domain prompt to choose the profile-generator path; a direct draft rewriter is a different tool.

A reusable profile deliverable is named

See the generated samples, analysis.md, and skill.md as named steps before reusing the profile.

A later-use domain boundary

Keep the profile tied to the chosen domain when it moves into later agent work.

A checkable output review

Review analysis.md, skill.md, and visual before-after evidence—or record the missing comparison—before handoff.

Start with a Domain

1Choose the domain whose writing or visual defaults you want the profile to capture.
2Install the Python CLI and run `python3 unslop.py --domain "blog writing"`; add `--type visual` when the domain is visual.
3Inspect the generated prompt set, samples, analysis.md, and skill.md; visual runs can also include screenshots and before-after comparisons.
4Review the profile’s domain boundary before using the generated Skill in another task.

Example Command

python3 unslop.py --domain "blog writing"

Core Capabilities

Domain prompt input

Start from a named domain; the CLI builds the prompt set and sample material for that domain.

Generated sample set

Produce text or visual samples, with screenshots available when the visual mode is selected.

analysis.md and skill.md

Keep the observed patterns in analysis.md before turning them into a reusable skill.md profile.

Before-and-after review

Review the generated profile and visual comparisons within the named domain instead of treating Unslop as a direct draft rewriter.

Use Cases

Brand voice study

Generate a profile for a brand-writing domain before using it in a later agent task.

Editorial pattern analysis

Give the publication domain to the CLI and review its generated patterns in analysis.md.

Reusable agent profile

Turn the generated domain analysis into a skill.md that can be reviewed and invoked later.

SKILL.md

# Unslop

This page is bound to Matt Shumer's MIT-licensed `mshumer/unslop` Skill wrapper at commit edcb62386d129c65e4395f0cfcc9168eb1ba2148. The source wraps a Python CLI that takes a domain, generates a prompt set and text or visual samples with Claude Code, and can render screenshots for visual work.

The job is generate → analyze the sample set → write analysis.md → generate a reusable skill.md profile → review the profile and comparisons. It is not a direct text rewriter, humanizer, or Stop Slop replacement. The same-keyword Unslop family contains other authors and jobs; this page is only Matt Shumer’s domain-profile generator.

HumanizerA separate source-preserving pass for reducing formulaic AI prose patterns.Stop SlopA separate red-pen editing workflow for preserving facts while revising prose.

Frequently Asked Questions

What is Matt Shumer’s Unslop Skill?

It is a domain-profile generator. Give its Python CLI a domain; it creates a prompt set and text or visual samples, writes analysis.md, and generates a reusable skill.md profile.

Is Unslop a direct text rewriter?

No. The selected source analyzes a domain and generates a reusable profile. It is not a one-click prose humanizer or direct draft rewrite.

How is this different from Stop Slop or Humanizer?

Stop Slop and Humanizer are separate published source entities focused on prose editing. Matt Shumer’s Unslop generates its own domain sample set and profile; do not merge their bodies or keyword demand.

How do I install Unslop, and what does it generate?

Clone mshumer/unslop, install its Python requirements, and run `python3 unslop.py --domain "blog writing"`; for visual work, add `--type visual`. The source generates a prompt set, samples, analysis.md, skill.md, and visual screenshots or before-after comparisons when requested.

How should I review the generated profile?

Compare analysis.md and skill.md with the generated samples and, for visual runs, the screenshots or before-after comparisons. Remove patterns that do not belong to the named domain before reuse.

Discussion

0 comments

Trigger Phrases

Use these phrases to activate this skill in your AI coding assistant:

Use Unslop to build a profile for this domain.Generate a reusable profile from a domain prompt.Analyze the generated samples before using the domain skill.