---

name: markdownlint-custom-rules description: Create custom linting rules for markdownlint including rule structure, parser integration, error reporting, and automatic fixing. allowed-tools: [Bash, Read]

Markdownlint Custom Rules

Master creating custom markdownlint rules including rule structure, markdown-it and micromark parser integration, error reporting with fixInfo, and asynchronous rule development.

Overview

Markdownlint allows you to create custom rules tailored to your project's specific documentation requirements. Custom rules can enforce project-specific conventions, validate content patterns, and ensure consistency beyond what built-in rules provide.

Rule Object Structure

Basic Rule Definition

Every custom rule must be a JavaScript object with specific properties:

module.exports = {
  names: ["rule-name", "RULE001"],
  description: "Description of what this rule checks",
  tags: ["custom", "style"],
  parser: "markdownit",
  function: function(params, onError) {
    // Rule implementation
  }
};

Required Properties

{
  names: Array<String>,        // Rule identifiers (required)
  description: String,         // What the rule checks (required)
  tags: Array<String>,         // Categorization tags (required)
  parser: String,              // "markdownit", "micromark", or "none" (required)
  function: Function          // Rule logic (required)
}

Optional Properties

{
  information: URL,           // Link to rule documentation
  asynchronous: Boolean      // If true, function returns Promise
}

Parser Selection

markdown-it Parser

Best for token-based parsing with rich metadata:

module.exports = {
  names: ["any-blockquote-markdown-it"],
  description: "Rule that reports an error for any blockquote",
  information: new URL("https://example.com/rules/any-blockquote"),
  tags: ["test"],
  parser: "markdownit",
  function: (params, onError) => {
    const blockquotes = params.parsers.markdownit.tokens
      .filter((token) => token.type === "blockquote_open");

    for (const blockquote of blockquotes) {
      const [startIndex, endIndex] = blockquote.map;
      const lines = endIndex - startIndex;

      onError({
        lineNumber: blockquote.lineNumber,
        detail: `Blockquote spans ${lines} line(s).`,
        context: blockquote.line
      });
    }
  }
};

micromark Parser

Best for detailed token analysis and precise positioning:

module.exports = {
  names: ["any-blockquote-micromark"],
  description: "Rule that reports an error for any blockquote",
  information: new URL("https://example.com/rules/any-blockquote"),
  tags: ["test"],
  parser: "micromark",
  function: (params, onError) => {
    const blockquotes = params.parsers.micromark.tokens
      .filter((token) => token.type === "blockQuote");

    for (const blockquote of blockquotes) {
      const lines = blockquote.endLine - blockquote.startLine + 1;

      onError({
        lineNumber: blockquote.startLine,
        detail: `Blockquote spans ${lines} line(s).`,
        context: params.lines[blockquote.startLine - 1]
      });
    }
  }
};

No Parser

For simple line-based rules:

module.exports = {
  names: ["no-todo-comments"],
  description: "Disallow TODO comments in markdown",
  tags: ["custom"],
  parser: "none",
  function: (params, onError) => {
    params.lines.forEach((line, index) => {
      if (line.includes("TODO:") || line.includes("FIXME:")) {
        onError({
          lineNumber: index + 1,
          detail: "TODO/FIXME comments should be resolved",
          context: line.trim()
        });
      }
    });
  }
};

Function Parameters

params Object

The params object contains all information about the markdown content:

function rule(params, onError) {
  // params.name - Input file/string name
  // params.lines - Array of lines (string[])
  // params.frontMatterLines - Lines of front matter
  // params.config - Rule's configuration from .markdownlint.json
  // params.version - markdownlint library version
  // params.parsers - Parser outputs
}

Accessing Lines

function: (params, onError) => {
  params.lines.forEach((line, index) => {
    const lineNumber = index + 1;  // Lines are 1-based

    if (someCondition(line)) {
      onError({
        lineNumber,
        detail: "Issue description",
        context: line.trim()
      });
    }
  });
}

Using Configuration

// In .markdownlint.json
{
  "custom-rule": {
    "max_length": 50,
    "pattern": "^[A-Z]"
  }
}

// In rule
function: (params, onError) => {
  const config = params.config || {};
  const maxLength = config.max_length || 40;
  const pattern = config.pattern ? new RegExp(config.pattern) : null;

  // Use configuration values
}

Working with Front Matter

function: (params, onError) => {
  const frontMatterLines = params.frontMatterLines;

  if (frontMatterLines.length > 0) {
    // Process YAML front matter
    const frontMatter = frontMatterLines.join('\n');
    // Validate front matter
  }
}

Error Reporting with onError

Basic Error Reporting

onError({
  lineNumber: 5,                          // Required: 1-based line number
  detail: "Line exceeds maximum length",  // Optional: Additional info
  context: "This is the problematic..."   // Optional: Relevant text
});

Error with Range

Highlight specific portion of the line:

onError({
  lineNumber: 10,
  detail: "Invalid heading format",
  context: "### Heading",
  range: [1, 3]  // Column 1, length 3 (highlights "###")
});

Error with Fix Information

Enable automatic fixing:

onError({
  lineNumber: 15,
  detail: "Extra whitespace",
  context: "  text  ",
  fixInfo: {
    editColumn: 1,
    deleteCount: 2,
    insertText: ""
  }
});

Automatic Fixing with fixInfo

Delete Characters

// Remove 5 characters starting at column 10
fixInfo: {
  lineNumber: 5,
  editColumn: 10,
  deleteCount: 5
}

Insert Text

// Insert text at column 1
fixInfo: {
  lineNumber: 3,
  editColumn: 1,
  insertText: "# "
}

Replace Text

// Replace 3 characters with new text
fixInfo: {
  lineNumber: 7,
  editColumn: 5,
  deleteCount: 3,
  insertText: "new"
}

Delete Entire Line

// Delete the entire line
fixInfo: {
  lineNumber: 10,
  deleteCount: -1
}

Insert New Line

// Insert a blank line
fixInfo: {
  lineNumber: 8,
  insertText: "\n"
}

Multi-Line Fix

Report multiple fixes for the same violation:

function: (params, onError) => {
  // Fix requires changes on multiple lines
  onError({
    lineNumber: 5,
    detail: "Inconsistent list markers",
    fixInfo: {
      lineNumber: 5,
      editColumn: 1,
      deleteCount: 1,
      insertText: "-"
    }
  });

  onError({
    lineNumber: 6,
    detail: "Inconsistent list markers",
    fixInfo: {
      lineNumber: 6,
      editColumn: 1,
      deleteCount: 1,
      insertText: "-"
    }
  });
}

Complete Rule Examples

Enforce Heading Capitalization

module.exports = {
  names: ["heading-capitalization", "HC001"],
  description: "Headings must start with a capital letter",
  tags: ["headings", "custom"],
  parser: "markdownit",
  function: (params, onError) => {
    const headings = params.parsers.markdownit.tokens
      .filter(token => token.type === "heading_open");

    for (const heading of headings) {
      const headingLine = params.lines[heading.lineNumber - 1];
      const match = headingLine.match(/^#+\s+(.+)$/);

      if (match) {
        const text = match[1];
        const firstChar = text.charAt(0);

        if (firstChar !== firstChar.toUpperCase()) {
          const hashCount = headingLine.indexOf(' ');

          onError({
            lineNumber: heading.lineNumber,
            detail: "Heading must start with capital letter",
            context: headingLine,
            range: [hashCount + 2, 1],
            fixInfo: {
              editColumn: hashCount + 2,
              deleteCount: 1,
              insertText: firstChar.toUpperCase()
            }
          });
        }
      }
    }
  }
};

Require Blank Line Before Headings

module.exports = {
  names: ["blank-line-before-heading", "BLH001"],
  description: "Require blank line before headings (except first line)",
  tags: ["headings", "custom", "whitespace"],
  parser: "markdownit",
  function: (params, onError) => {
    const headings = params.parsers.markdownit.tokens
      .filter(token => token.type === "heading_open");

    for (const heading of headings) {
      const lineNumber = heading.lineNumber;

      // Skip if first line or after front matter
      if (lineNumber <= params.frontMatterLines.length + 1) {
        continue;
      }

      const previousLine = params.lines[lineNumber - 2];

      if (previousLine.trim() !== "") {
        onError({
          lineNumber: lineNumber - 1,
          detail: "Expected blank line before heading",
          context: previousLine,
          fixInfo: {
            lineNumber: lineNumber - 1,
            editColumn: previousLine.length + 1,
            insertText: "\n"
          }
        });
      }
    }
  }
};

Validate Code Block Language

module.exports = {
  names: ["code-block-language", "CBL001"],
  description: "Code blocks must specify a language",
  tags: ["code", "custom"],
  parser: "markdownit",
  function: (params, onError) => {
    const config = params.config || {};
    const allowedLanguages = config.allowed_languages || [];

    const fences = params.parsers.markdownit.tokens
      .filter(token => token.type === "fence");

    for (const fence of fences) {
      const language = fence.info.trim();

      if (!language) {
        onError({
          lineNumber: fence.lineNumber,
          detail: "Code block must specify a language",
          context: fence.line
        });
      } else if (allowedLanguages.length > 0 && !allowedLanguages.includes(language)) {
        onError({
          lineNumber: fence.lineNumber,
          detail: `Language '${language}' not in allowed list: ${allowedLanguages.join(', ')}`,
          context: fence.line
        });
      }
    }
  }
};

Detect Broken Relative Links

const fs = require('fs');
const path = require('path');

module.exports = {
  names: ["no-broken-links", "NBL001"],
  description: "Detect broken relative links",
  tags: ["links", "custom"],
  parser: "markdownit",
  asynchronous: true,
  function: async (params, onError) => {
    const links = params.parsers.markdownit.tokens
      .filter(token => token.type === "link_open");

    for (const link of links) {
      const hrefToken = link.attrs.find(attr => attr[0] === "href");

      if (hrefToken) {
        const href = hrefToken[1];

        // Only check relative links
        if (!href.startsWith('http://') && !href.startsWith('https://')) {
          const filePath = path.join(path.dirname(params.name), href);

          try {
            await fs.promises.access(filePath);
          } catch (err) {
            onError({
              lineNumber: link.lineNumber,
              detail: `Broken link: ${href}`,
              context: link.line
            });
          }
        }
      }
    }
  }
};

Enforce Consistent List Markers

module.exports = {
  names: ["consistent-list-markers", "CLM001"],
  description: "Lists must use consistent markers within the same level",
  tags: ["lists", "custom"],
  parser: "micromark",
  function: (params, onError) => {
    const lists = params.parsers.micromark.tokens
      .filter(token => token.type === "listUnordered");

    for (const list of lists) {
      const items = params.parsers.micromark.tokens.filter(
        token => token.type === "listItemMarker" &&
                 token.startLine >= list.startLine &&
                 token.endLine <= list.endLine
      );

      if (items.length > 0) {
        const firstMarker = params.lines[items[0].startLine - 1]
          .charAt(items[0].startColumn - 1);

        for (const item of items.slice(1)) {
          const marker = params.lines[item.startLine - 1]
            .charAt(item.startColumn - 1);

          if (marker !== firstMarker) {
            onError({
              lineNumber: item.startLine,
              detail: `Inconsistent list marker: expected '${firstMarker}', found '${marker}'`,
              context: params.lines[item.startLine - 1],
              range: [item.startColumn, 1],
              fixInfo: {
                editColumn: item.startColumn,
                deleteCount: 1,
                insertText: firstMarker
              }
            });
          }
        }
      }
    }
  }
};

Asynchronous Rules

Basic Async Rule

module.exports = {
  names: ["async-rule-example"],
  description: "Example asynchronous rule",
  tags: ["async", "custom"],
  parser: "none",
  asynchronous: true,
  function: async (params, onError) => {
    // Can use await
    const result = await someAsyncOperation();

    if (!result.valid) {
      onError({
        lineNumber: 1,
        detail: "Async validation failed"
      });
    }

    // Must return Promise (implicitly returned by async function)
  }
};

Network Validation

const https = require('https');

module.exports = {
  names: ["validate-external-links"],
  description: "Validate external HTTP links return 200",
  tags: ["links", "async"],
  parser: "markdownit",
  asynchronous: true,
  function: async (params, onError) => {
    const links = params.parsers.markdownit.tokens
      .filter(token => token.type === "link_open");

    const checkLink = (url) => {
      return new Promise((resolve) => {
        https.get(url, (res) => {
          resolve(res.statusCode === 200);
        }).on('error', () => {
          resolve(false);
        });
      });
    };

    for (const link of links) {
      const hrefToken = link.attrs.find(attr => attr[0] === "href");

      if (hrefToken) {
        const href = hrefToken[1];

        if (href.startsWith('http://') || href.startsWith('https://')) {
          const valid = await checkLink(href);

          if (!valid) {
            onError({
              lineNumber: link.lineNumber,
              detail: `External link may be broken: ${href}`,
              context: link.line
            });
          }
        }
      }
    }
  }
};

Using Custom Rules

In Configuration File

// .markdownlint.js
const customRules = require('./custom-rules');

module.exports = {
  default: true,
  customRules: [
    customRules.headingCapitalization,
    customRules.blankLineBeforeHeading,
    customRules.codeBlockLanguage
  ],
  "heading-capitalization": true,
  "blank-line-before-heading": true,
  "code-block-language": {
    "allowed_languages": ["javascript", "typescript", "bash", "json"]
  }
};

In Node.js Script

const markdownlint = require('markdownlint');
const customRules = require('./custom-rules');

const options = {
  files: ['README.md'],
  customRules: [
    customRules.headingCapitalization,
    customRules.blankLineBeforeHeading
  ],
  config: {
    default: true,
    "heading-capitalization": true,
    "blank-line-before-heading": true
  }
};

markdownlint(options, (err, result) => {
  if (!err) {
    console.log(result.toString());
  }
});

With markdownlint-cli

# Using custom rules with CLI
markdownlint -c .markdownlint.js -r ./custom-rules/*.js *.md

TypeScript Support

Type-Safe Rule Definition

import { Rule } from 'markdownlint';

const rule: Rule = {
  names: ['typescript-rule', 'TS001'],
  description: 'Example TypeScript custom rule',
  tags: ['custom'],
  parser: 'markdownit',
  function: (params, onError) => {
    // Type-safe implementation
    params.parsers.markdownit.tokens.forEach(token => {
      if (token.type === 'heading_open') {
        onError({
          lineNumber: token.lineNumber,
          detail: 'Example error'
        });
      }
    });
  }
};

export default rule;

When to Use This Skill

• Enforcing project-specific documentation standards
• Validating custom markdown patterns
• Checking domain-specific requirements
• Extending markdownlint beyond built-in rules
• Creating reusable rule packages
• Automating documentation quality checks
• Implementing team coding standards
• Building custom linting toolchains

Best Practices

1. Clear Rule Names - Use descriptive names that indicate purpose
2. Comprehensive Descriptions - Document what the rule checks
3. Appropriate Tags - Categorize rules for easy filtering
4. Choose Right Parser - Use markdownit for most cases, micromark for precision
5. Provide Information URLs - Link to detailed rule documentation
6. Support Configuration - Allow rule customization via params.config
7. Helpful Error Messages - Provide clear detail and context
8. Use Range When Possible - Highlight exact problem location
9. Implement fixInfo - Enable automatic fixing when possible
10. Handle Edge Cases - Account for front matter, empty files, etc.
11. Performance Consideration - Avoid expensive operations in rules
12. Test Thoroughly - Test with various markdown files
13. Version Documentation - Document which markdownlint version required
14. Export Properly - Use module.exports or ES6 exports consistently
15. Async When Needed - Only use asynchronous for I/O operations

Common Pitfalls

1. Wrong Line Numbers - Forgetting lines are 1-based, not 0-based
2. Missing Parser - Not specifying parser property
3. Incorrect Token Types - Using wrong token type names
4. No Error Context - Not providing helpful context in errors
5. Synchronous I/O - Using sync functions instead of async
6. Ignoring Front Matter - Not handling front matter correctly
7. Hardcoded Values - Not using configuration parameters
8. Poor Performance - Using inefficient algorithms on large files
9. Missing Fixability - Not implementing fixInfo when possible
10. Incomplete Testing - Not testing edge cases and error conditions
11. Parser Mismatch - Accessing wrong parser output
12. Column Off-by-One - Columns are 1-based like line numbers
13. Memory Leaks - Not cleaning up in async rules
14. Blocking Operations - Long-running synchronous operations
15. Type Confusion - Mixing up token properties between parsers

Resources

• markdownlint Custom Rules Guide
• markdownlint GitHub Repository
• markdown-it Parser
• micromark Parser
• Community Custom Rules on npm