Skip to content

Zero Slop skill

Edit drafts with an existing AI assistant and local writing checks that preserve source details.

Community · Manav Mishra
AuthorManav Mishra
LicenceMIT
Sourcemanavmishra/ZeroSlop · SKILL.md on GitHub
Project websitezero-slop.ai
Works inClaude Code · Codex, Gemini CLI, Cursor, Copilot
Read from the source on

Zero Slop combines an editorial workflow with local Python checks for wording, rhythm and source-detail preservation.

The existing assistant performs the reading and rewriting. Eight roles describe responsibilities, not eight independent models; a clean writing score does not prove authorship or factual safety.

From the skill's SKILL.md: # Zero Slop

This is a community skill by Manav Mishra, not checked by us or by Anthropic. Skills can tell the agent to run code, read files or call the network. Before installing, read its SKILL.md and every script in the folder, and look for anything that does not match what it says it does.

What it does

  1. Score the draft locally and review context beyond lexical patterns.
  2. Edit, copyedit and read the result against the supplied source.
  3. Check the final version and disclose residual concerns or unavailable tools.

What you get

  • An edited draft with plain-language verification notes.
  • Optional inspect-only findings or simulated audience-review notes.

When to use it

  • Polish outward-facing prose while retaining claims and voice.
  • Request an explicit audience simulation without altering the draft.

When not to use it

  • Treating simulated readers as real feedback or promising detector bypass.

What it needs

  • An existing AI assistant; Python 3 for bundled local scoring and preservation tools.
  • The optional version check contacts GitHub for release metadata; ZS_NO_UPDATE_CHECK=1 disables it.
  • Hosted MCP, REST and npm deslop are separate remote services that receive drafts; local skill checks do not require them.

Files in the skill

File or folderWhat it is for
SKILL.mdEditorial responsibilities and boundaries.
references/detailed-workflow.mdRequired execution and reporting sequence.
scripts/slopscore.pyLocal writing-pattern scorer.
scripts/reader_review.pyLocal passage packets and review-page preparation.

Licence as found in the repository on 2026-11-17: MIT.

Example prompts

Type one of these after installing; naming the skill makes sure it is used.

  • Edit this email using local checks only; preserve all names, numbers and caveats.
  • Inspect the wording without rewriting and distinguish measured signals from judgment.
  • Simulate two specified reader perspectives, leave the draft unchanged and label the reactions as hypotheses.

How to install zero-slop

Local and remote paths

Install the complete skills/zero-slop folder with its scripts, references and data. Python checks run locally; the host assistant still processes the draft under its own terms.

Do not confuse local scoring with npm deslop or hosted MCP/REST rewriting. Those optional routes send drafts to a remote service and have separate availability and privacy terms.

A skill is a folder with a SKILL.md file. You install it once; the agent reads its short description at the start of each session and loads the full instructions only when a request matches. Pick the place you use:

Claude Code: copy the folder into your skills folder

Put the skill's folder in ~/.claude/skills/ to use it in every project on this computer, or in .claude/skills/ inside a repository to share it with everyone who works there (commit it).

git clone --depth 1 https://github.com/manavmishra/ZeroSlop.git
mkdir -p ~/.claude/skills
cp -R ZeroSlop/skills/zero-slop ~/.claude/skills/

Claude Code notices new or changed skills in these folders during the session. If the skills folder did not exist when the session started, run /reload-skills once. Then type /zero-slop or just describe the task.

On Windows, run the commands in Git Bash or WSL, or copy the folder into the same place with File Explorer.

Codex, Gemini CLI, Cursor and GitHub Copilot

These agents read the same SKILL.md folders. The one folder all four accept inside a project is .agents/skills/, so this copies the skill there:

git clone --depth 1 https://github.com/manavmishra/ZeroSlop.git
mkdir -p .agents/skills
cp -R ZeroSlop/skills/zero-slop .agents/skills/
AgentProject folderYour user folderNotes
OpenAI Codex
.agents/skills/
~/.agents/skills/
Type $ or run /skills to pick a skill; Codex also chooses one when your task matches.
Gemini CLI
.agents/skills/
.gemini/skills/
~/.agents/skills/
~/.gemini/skills/
Or install from Git with gemini skills install <git URL> --path <folder> --consent. Google replaced Gemini CLI with Antigravity CLI for free-tier and Google One users on 18 June 2026; Antigravity uses .agents/skills/ in a project and ~/.gemini/config/skills/ for you.
Cursor
.agents/skills/
.cursor/skills/
~/.agents/skills/
~/.cursor/skills/
Type / in Agent chat to pick a skill. Cursor also reads Claude Code's .claude/skills/ folders.
GitHub Copilot
.github/skills/
.agents/skills/
.claude/skills/
~/.copilot/skills/
~/.agents/skills/
Works in Copilot CLI, the cloud agent, and agent mode in VS Code, Visual Studio and JetBrains (IDE support depends on the version). Or run gh skill install OWNER/REPOSITORY SKILL (public preview, GitHub CLI 2.90.0 or later).
Only install skills you trust. Anthropic's advice is to use skills only from trusted sources, and to read every file of any other skill before using it: a skill can direct the agent to run commands, read files or send data in ways its description does not mention.

Sources for these steps

Checked against each vendor's documentation on 2026-11-14:

Checks have limits

Source-detail checks cannot prove that a reframed claim kept its meaning. Compare the exact final draft, including qualifiers and speaker perspective.

Optional character cleanup is conservative and reversible. It does not prove removal of statistical watermarks or replace contextual editing.

Questions about zero-slop

Is Zero Slop another AI model?

No. Your current assistant performs contextual editing; bundled tools handle repeatable local checks.

Are eight roles eight independent reviews?

No. They are ordered responsibilities; one model response is not independent verification.

Does a low score prove the draft is human?

No. Scores measure tracked writing patterns and cannot identify an author.

Does the reader-review helper call a model?

No. It prepares passage packets and a local page. The assistant supplies simulated reactions, which are hypotheses rather than observed reader behavior.

More AI skills

Related skills

Skill categories