This skill makes an AI agent write in a plain style. The full rules are in
skills/plain-writing/SKILL.md.
The rules are plain text, so any agent can use them.
The skill also checks its own writing and removes anything that does not add something.
skills/plain-writing/SKILL.md: the skill, with the rules and the steps.evals/: optional. Only needed if you want to measure the skill.
skills/plain-writing/SKILL.md is a plain markdown file with the rules. Any
agent that can read a file can follow it. The simplest way is to give that
file to the agent as instructions, in a rules file or the system prompt.
The skill lives under skills/ so gh skill install can find it:
gh skill install docwriter-org/plain-writing-skill plain-writing
Some tools have a set place for skills:
- Claude Code reads skills from
~/.claude/skills. Install only the skill file, not the evals:
mkdir -p ~/.claude/skills/plain-writing
curl -fsSL https://raw.githubusercontent.com/docwriter-org/plain-writing-skill/main/skills/plain-writing/SKILL.md \
-o ~/.claude/skills/plain-writing/SKILL.md
- Other agents, e.g., Codex or pi, can use the rules too. Paste the rules from
skills/plain-writing/SKILL.mdinto whatever instructions that agent reads.
Then ask the agent to write or revise some text. It applies the rules on its own.
Use the command below when an agent response assumes too much context, uses unexplained technical terms, or does not make the main point clear.
/plain-writing deslopify
The skill rewrites the agent's previous response in a clear structure for a sharp CEO or technical reader who has no project context and needs to understand all relevant details. The rewrite follows the plain-writing guidelines in the skill.
You can also put text after the command when you want to rewrite text other than the previous response.
A skill install does not need the evals. If you want to run them, clone the
repo and see evals/README.md.