Skip to content
Back to skills

Md Writer

ASecurity

Write or edit knowledge-base Markdown with frontmatter, Mermaid diagrams, related-document links, and markdownlint validation. Use for guides, specs, ADRs, runbooks, and API documentation; not repository meta files such as README, CHANGELOG, CLAUDE.md, AGENTS.md, or SKILL.md.

  • 2 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added October 22, 2026
ai-agentsgobashapidocumentation

Works with

  • api

Security analysis

A100/100

Pro scans all 9 files and shows the line behind each finding

Scanned October 22, 2026

npx -y skills add skrrt-sh/skills --skill md-writer --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Md Writer?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Md Writer
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/skrrt-sh-md-writer/badge)](https://www.skillsdirectory.com/skills/skrrt-sh-md-writer)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: md-writer
description: Write or edit knowledge-base Markdown with frontmatter, Mermaid diagrams, related-document links, and markdownlint validation. Use for guides, specs, ADRs, runbooks, and API documentation; not repository meta files such as README, CHANGELOG, CLAUDE.md, AGENTS.md, or SKILL.md.
allowed-tools: Bash(${CLAUDE_SKILL_DIR}/scripts/validate-md.sh:*)
---

# Markdown Authoring

Write readable, lint-clean knowledge-base documents. Match an established local convention before
applying this skill's defaults.

## Document shape

Use this frontmatter for new documents, omitting fields whose values are unknown:

```yaml
---
title: "Document title"
description: "One-sentence purpose"
created: "YYYY-MM-DD"
updated: "YYYY-MM-DD"
status: "draft"
---
```

Add `author`, `version`, `tags`, `category`, `aliases`, `related`, `refs`, or `audience` only when
the user, repository, or document supplies a real value. Never invent metadata or body content;
leave unknown sections pending or ask. For new documents, use today's date and
`version: "1.0.0"` only when the repository versions individual documents.

Make the H1 match `title`. Follow it with a one-line blockquote summary. Add a table of contents
when it materially improves navigation, typically at five or more H2 sections.

## Diagrams

Use Mermaid instead of ASCII art. Choose the diagram type that fits the relationship. Use `<br/>`
for line breaks in labels; a literal `\n` does not render as a break.

```mermaid
flowchart TD
    A["First line<br/>second line"] --> B["Result"]
```

## Formatting

- Wrap prose at 120 characters; exempt code blocks and tables.
- Avoid inline HTML outside Mermaid content.
- Use compact tables with single-space cell padding.
- Name files with lowercase hyphenated words.
- Give every fenced code block a language.

## Related documents

For a document in an existing knowledge base, search nearby Markdown files and titles for genuine
topic, dependency, or parent/child overlap. Add deduplicated relative paths to this document's
`related` field. Edit existing documents to add backlinks only when the user asked for
bidirectional links; update only their `related` and `updated` fields.

## Validate

From the target repository, run
`${CLAUDE_SKILL_DIR}/scripts/validate-md.sh <file.md>` on every changed in-scope Markdown
file. It may rewrite mechanical formatting; re-read a changed file before editing it again. Fix
reported judgment-based violations and repeat until clean.

Read [references/validation.md](references/validation.md) only when validation is skipped, cannot
start, or reports a violation.

Files in this skill

  • SKILL.md2.6 KB
  • agents/openai.yaml244 B
  • config/markdownlint-default.json242 B
  • evals/evals.json5.7 KB
  • evals/trigger-evals.json2.1 KB
  • package-lock.json45.6 KB
  • package.json463 B
  • references/validation.md1.2 KB
  • scripts/validate-md.sh4.2 KB

Attribution

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

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.