Learn-skills.dev style-prose
install
source · Clone the upstream repo
git clone https://github.com/NeverSight/learn-skills.dev
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/NeverSight/learn-skills.dev "$T" && mkdir -p ~/.claude/skills && cp -r "$T/data/skills-md/agglayer/agglayer-ai-skills/style-prose" ~/.claude/skills/neversight-learn-skills-dev-style-prose && rm -rf "$T"
manifest:
data/skills-md/agglayer/agglayer-ai-skills/style-prose/SKILL.mdsource content
When writing or editing text files (markdown, plain text, etc.), use Semantic Line Breaks (sembr.org).
Rules
- Line length: aim for 80-100 characters per line. This is the primary constraint; the rules below serve it.
- Add a line break after each sentence (after
,.
, or!
).? - Break at a clause boundary when a line would otherwise exceed ~100 characters.
- Do not micro-split: short clauses, enumerations, and closely related phrases should stay on the same line when they fit within the target length.
- Break before enumerated or itemized lists.
- Do not break within hyphenated words.
- Allow longer lines for URLs, code spans, or markup.
Why
Semantic line breaks make diffs cleaner: editing one clause changes only one line, not the entire paragraph. The rendered output is unaffected.
Scope
Apply to all prose in markup files (markdown, rst, adoc, plain text). Do not apply to code blocks, YAML frontmatter, or structured data inside these files.