diff --git a/src/index.d.ts b/src/index.d.ts
deleted file mode 100644
index 9e1191f..0000000
--- a/src/index.d.ts
+++ /dev/null
@@ -1,111 +0,0 @@
-/**
- * Matches markdown headers (lines starting with one or more # characters).
- * Example: `# Heading 1`, `## Heading 2`
- */
-export const REGEXP_HEADER: RegExp;
-
-/**
- * Matches markdown level 2 headers (lines starting with exactly ##).
- * Example: `## Heading 2`
- */
-export const REGEXP_H2: RegExp;
-
-/**
- * Matches markdown level 3 headers (lines starting with exactly ###).
- * Example: `### Heading 3`
- */
-export const REGEXP_H3: RegExp;
-
-/**
- * Matches markdown image syntax.
- * Example: `![alt text](https://example.com/image.png)`
- */
-export const REGEXP_IMAGE: RegExp;
-
-/**
- * Matches markdown link syntax.
- * Example: `[link text](https://example.com)`
- */
-export const REGEXP_LINK: RegExp;
-
-/**
- * Matches bold/strong text using `**` or `__` delimiters.
- * Example: `**bold**`, `__bold__`
- */
-export const REGEXP_STRONG: RegExp;
-
-/**
- * Matches strikethrough text using `~~` delimiters.
- * Example: `~~strikethrough~~`
- */
-export const REGEXP_DEL: RegExp;
-
-/**
- * Matches quoted text using `::"...":` syntax.
- * Example: `::"quoted text":`
- */
-export const REGEXP_Q: RegExp;
-
-/**
- * Matches inline code using backtick delimiters.
- * Example: `` `code` ``
- */
-export const REGEXP_CODE: RegExp;
-
-/**
- * Matches unordered list items (lines starting with `*`).
- * Example: `* list item`
- */
-export const REGEXP_UL_LIST: RegExp;
-
-/**
- * Matches ordered list items (lines starting with a number and period).
- * Example: `1. list item`
- */
-export const REGEXP_OL_LIST: RegExp;
-
-/**
- * Matches blockquote lines starting with `>` or `>`.
- * Example: `> blockquote text`
- */
-export const REGEXP_BLOCKQUOTE: RegExp;
-
-/**
- * Matches horizontal rule lines (5 or more dashes).
- * Example: `-----`
- */
-export const REGEXP_HR: RegExp;
-
-/**
- * Matches paragraph content between newlines.
- */
-export const REGEXP_PARAGRAPH: RegExp;
-
-/**
- * Matches empty unordered list HTML tags (consecutive `</ul><ul>` pairs).
- * Example: `</ul><ul>`
- */
-export const REGEXP_EMPTY_UL: RegExp;
-
-/**
- * Matches empty ordered list HTML tags (consecutive `</ol><ol>` pairs).
- * Example: `</ol><ol>`
- */
-export const REGEXP_EMPTY_OL: RegExp;
-
-/**
- * Matches two or more consecutive newlines (line breaks).
- */
-export const REGEXP_BR: RegExp;
-
-/**
- * Matches consecutive empty blockquote HTML tags.
- * Example: `</blockquote><blockquote>`
- */
-export const REGEXP_EMPTY_BLOCKQUOTE: RegExp;
-
-/**
- * Matches italic/emphasis text using `*` or `_` delimiters.
- * Example: `*italic*`, `_italic_`
- */
-export const REGEXP_EM: RegExp;
diff --git a/src/index.js b/src/index.js
deleted file mode 100644
index cb39867..0000000
--- a/src/index.js
+++ /dev/null
@@ -1,27 +0,0 @@
-/**
- * Legacy CommonJS/ESM entry kept for compatibility during migration.
- * Prefer the TypeScript source (src/index.ts) and the built dist/ package.
- */
-export {
-  REGEXP_HEADER,
-  REGEXP_H2,
-  REGEXP_H3,
-  REGEXP_IMAGE,
-  REGEXP_LINK,
-  REGEXP_STRONG,
-  REGEXP_DEL,
-  REGEXP_Q,
-  REGEXP_CODE,
-  REGEXP_BLOCKQUOTE,
-  REGEXP_HR,
-  REGEXP_PARAGRAPH,
-  REGEXP_BR,
-  REGEXP_EMPTY_BLOCKQUOTE,
-  REGEXP_ITALIC,
-} from './tags/index.js';
-export {
-  REGEXP_UL_LIST,
-  REGEXP_OL_LIST,
-  REGEXP_EMPTY_UL,
-  REGEXP_EMPTY_OL,
-} from './lists/index.js';
diff --git a/src/index.ts b/src/index.ts
index 6c4bba2..76b2762 100644
--- a/src/index.ts
+++ b/src/index.ts
@@ -1,10 +1,10 @@
 /**
  * markdown-regex
  *
- * A set of ready-to-use RegExp constants for parsing and extracting
- * elements from Markdown content.
- *
+ * Ready-to-use RegExp constants for parsing Markdown.
  * Zero runtime dependencies. Works in Node, browsers, and bundlers.
+ *
+ * @packageDocumentation
  */
 
 export {
diff --git a/src/list.js b/src/list.js
deleted file mode 100644
index 24ce169..0000000
--- a/src/list.js
+++ /dev/null
@@ -1,23 +0,0 @@
-import os from "os";
-
-const platform = os.platform();
-
-const newLine = platform === "win32" ? "\r\n" : "\n";
-
-const REGEXP_UL_LIST = new RegExp(
-  `${newLine}(((\\s{4})?\\*(.*?)${newLine}){1,})`,
-  "g"
-);
-const REGEXP_OL_LIST = new RegExp(`${newLine}[0-9]+\\.(.*)`, "g");
-
-// @TODO make all regEx the same style
-const REGEXP_EMPTY_UL = /<\/ul>\s?<ul>/g;
-const REGEXP_EMPTY_OL = /<\/ol>\s?<ol>/g;
-
-export {
-  REGEXP_UL_LIST,
-  REGEXP_OL_LIST,
-  // -----------
-  REGEXP_EMPTY_UL,
-  REGEXP_EMPTY_OL,
-};
diff --git a/src/lists/empty-ol.ts b/src/lists/empty-ol.ts
index 45d7a49..c1333ac 100644
--- a/src/lists/empty-ol.ts
+++ b/src/lists/empty-ol.ts
@@ -1,7 +1,8 @@
 /**
- * Matches consecutive ordered list HTML tags with optional whitespace.
- * Useful for cleaning up generated HTML.
+ * Matches consecutive empty ordered-list HTML tags.
+ * Useful when cleaning generated HTML.
  *
- * Example: </ol><ol> or </ol> <ol>
+ * @example </ol><ol>
+ * @example </ol> <ol>
  */
 export const REGEXP_EMPTY_OL = /<\/ol>\s?<ol>/g;
diff --git a/src/lists/empty-ul.ts b/src/lists/empty-ul.ts
index 85f5280..54aa3e0 100644
--- a/src/lists/empty-ul.ts
+++ b/src/lists/empty-ul.ts
@@ -1,7 +1,8 @@
 /**
- * Matches consecutive unordered list HTML tags with optional whitespace.
- * Useful for cleaning up generated HTML.
+ * Matches consecutive empty unordered-list HTML tags.
+ * Useful when cleaning generated HTML.
  *
- * Example: </ul><ul> or </ul> <ul>
+ * @example </ul><ul>
+ * @example </ul> <ul>
  */
 export const REGEXP_EMPTY_UL = /<\/ul>\s?<ul>/g;
diff --git a/src/lists/ol.ts b/src/lists/ol.ts
index 9c18748..90168dd 100644
--- a/src/lists/ol.ts
+++ b/src/lists/ol.ts
@@ -1,4 +1,6 @@
-import { NEWLINE } from '../utils';
-
-/** Matches ordered list items: 1. item */
-export const REGEXP_OL_LIST = new RegExp(NEWLINE + '[0-9]+\\.(.*)', 'g');
+/**
+ * Matches an ordered list item (number + period after a newline).
+ *
+ * @example 1. first item
+ */
+export const REGEXP_OL_LIST = /(?:\r\n|\r|\n)\d+\.(.*)/g;
diff --git a/src/lists/ul.ts b/src/lists/ul.ts
index 800189c..d122d58 100644
--- a/src/lists/ul.ts
+++ b/src/lists/ul.ts
@@ -1,15 +1,9 @@
-import { NEWLINE } from '../utils';
-
 /**
- * Matches unordered list items (lines starting with *).
- * Supports optional leading indentation (4 spaces).
- * Platform-agnostic newline handling.
+ * Matches an unordered list item (`*` after a newline).
+ * Optional 4-space indent is allowed.
  *
- * Example:
- * * item one
- * * item two
+ * @example
+ *   * item one
+ *   * item two
  */
-export const REGEXP_UL_LIST = new RegExp(
-  `${NEWLINE}(((?:\\s{4})?\\*(.*?)${NEWLINE}){1,})`,
-  'g'
-);
+export const REGEXP_UL_LIST = /(?:\r\n|\r|\n)(?:\s{4})?\*(.*)/g;
diff --git a/src/tags.js b/src/tags.js
deleted file mode 100644
index 6d3ed7d..0000000
--- a/src/tags.js
+++ /dev/null
@@ -1,54 +0,0 @@
-import os from "os";
-
-const platform = os.platform();
-
-const newLine = platform === "win32" ? "\r\n" : "\n";
-
-// TODO: make all regex to feel similar
-const REGEXP_HEADER = new RegExp(`${newLine}(#+)(.*)`, "g");
-
-const REGEXP_H3 = /^### (.*$)/gim;
-const REGEXP_H2 = /^## (.*$)/gim;
-// -----------
-
-// TODO: make all regex to feel similar
-const REGEXP_IMAGE = /!\[([^\[]+)\]\(([^\)]+)\)/g;
-// TODO ADD REF PARAMETER TO URL
-const REGEXP_LINK = /\[([^\[]+)\]\(([^\)]+)\)/g;
-const REGEXP_STRONG = /(\*\*|__)(.*?)(\*?)\1/g;
-const REGEXP_DEL = /\~\~(.*?)\~\~/g;
-const REGEXP_Q = /\:\"(.*?)\"\:/g;
-const REGEXP_CODE = /`(.*?)`/g;
-
-// -----------
-const REGEXP_BLOCKQUOTE = new RegExp(`${newLine}(&gt;|\\>)(.*)`, "g");
-const REGEXP_HR = new RegExp(`${newLine}-{5,}`, "g");
-
-const REGEXP_PARAGRAPH = new RegExp(`${newLine}(.+?)${newLine}`, "g");
-
-// -----------
-const REGEXP_BR = new RegExp(`((${newLine}){2,})`, "g");
-
-const REGEXP_EMPTY_BLOCKQUOTE = /<\/blockquote><blockquote>/g;
-
-const REGEXP_EM = /(\s|>)(\*|_)(.*?)\2(\s|<)/g;
-
-export {
-  REGEXP_HEADER,
-  REGEXP_H2,
-  REGEXP_H3,
-  REGEXP_IMAGE,
-  REGEXP_LINK,
-  REGEXP_STRONG,
-  REGEXP_DEL,
-  REGEXP_Q,
-  REGEXP_CODE,
-  // -----------
-  REGEXP_BLOCKQUOTE,
-  REGEXP_HR,
-  REGEXP_PARAGRAPH,
-  // -----------
-  REGEXP_BR,
-  REGEXP_EMPTY_BLOCKQUOTE,
-  REGEXP_EM,
-};
diff --git a/src/tags/blockquote.ts b/src/tags/blockquote.ts
index 7ddc951..75a0693 100644
--- a/src/tags/blockquote.ts
+++ b/src/tags/blockquote.ts
@@ -1,9 +1,7 @@
-import { NEWLINE } from '../utils';
-
 /**
- * Matches blockquote lines starting with > or &gt;.
- * Platform-agnostic newline handling.
+ * Matches blockquote lines starting with `>` or `&gt;` after a newline.
  *
- * Example: > blockquote text
+ * @example
+ *   > quoted paragraph
  */
-export const REGEXP_BLOCKQUOTE = new RegExp(`${NEWLINE}(&gt;|>)(.*)`, 'g');
+export const REGEXP_BLOCKQUOTE = /(?:\r\n|\r|\n)(?:>|&gt;)(.*)/g;
diff --git a/src/tags/br.ts b/src/tags/br.ts
index 44b094f..ebdb338 100644
--- a/src/tags/br.ts
+++ b/src/tags/br.ts
@@ -1,7 +1,7 @@
-import { NEWLINE } from '../utils';
-
 /**
  * Matches two or more consecutive newlines (paragraph / line breaks).
- * Platform-agnostic.
+ * Understands `\n`, `\r\n` and `\r`.
+ *
+ * @example a\\n\\nb
  */
-export const REGEXP_BR = new RegExp(`((${NEWLINE}){2,})`, 'g');
+export const REGEXP_BR = /(?:\r\n|\r|\n){2,}/g;
diff --git a/src/tags/code.ts b/src/tags/code.ts
index 4ebe11c..994b652 100644
--- a/src/tags/code.ts
+++ b/src/tags/code.ts
@@ -1,6 +1,6 @@
 /**
  * Matches inline code using single backtick delimiters.
  *
- * Example: `code`
+ * @example `code`
  */
-export const REGEXP_CODE = /`(.*?)`/g;
+export const REGEXP_CODE = /`(.+?)`/g;
diff --git a/src/tags/del.ts b/src/tags/del.ts
index 044d80d..f96b631 100644
--- a/src/tags/del.ts
+++ b/src/tags/del.ts
@@ -1,6 +1,6 @@
 /**
- * Matches strikethrough text using ~~ delimiters.
+ * Matches strikethrough text using `~~` delimiters.
  *
- * Example: ~~strikethrough~~
+ * @example ~~deleted~~
  */
-export const REGEXP_DEL = /~~(.*?)~~/g;
+export const REGEXP_DEL = /~~(.+?)~~/g;
diff --git a/src/tags/em.js b/src/tags/em.js
deleted file mode 100644
index 90999a5..0000000
--- a/src/tags/em.js
+++ /dev/null
@@ -1,4 +0,0 @@
-/**
- * @deprecated Use REGEXP_ITALIC from './italic.js'. Delete this file.
- */
-export { REGEXP_ITALIC as REGEXP_ITALIC_ONLY } from './italic.js';
diff --git a/src/tags/em.ts b/src/tags/em.ts
deleted file mode 100644
index 6f5927f..0000000
--- a/src/tags/em.ts
+++ /dev/null
@@ -1,6 +0,0 @@
-/**
- * @deprecated Removed. Use REGEXP_ITALIC from './italic' instead.
- * This file only exists so accidental imports fail clearly at review time.
- * Delete this file after the hygiene pass (`git rm src/tags/em.ts`).
- */
-export { REGEXP_ITALIC as REGEXP_ITALIC_ONLY } from './italic';
diff --git a/src/tags/empty-blockquote.ts b/src/tags/empty-blockquote.ts
index 0757740..1fc4125 100644
--- a/src/tags/empty-blockquote.ts
+++ b/src/tags/empty-blockquote.ts
@@ -1,7 +1,7 @@
 /**
  * Matches consecutive empty blockquote HTML tags.
- * Useful for cleaning up generated HTML.
+ * Useful when cleaning generated HTML.
  *
- * Example: </blockquote><blockquote>
+ * @example </blockquote><blockquote>
  */
 export const REGEXP_EMPTY_BLOCKQUOTE = /<\/blockquote><blockquote>/g;
diff --git a/src/tags/header.ts b/src/tags/header.ts
index 46c8164..985935a 100644
--- a/src/tags/header.ts
+++ b/src/tags/header.ts
@@ -1,6 +1,26 @@
-/** ATX headers after start-of-string or a newline. */
+/**
+ * Matches ATX markdown headers (lines starting with 1–6 `#` characters).
+ * Matches at the start of the string or after a newline (Unix / Windows / old Mac).
+ *
+ * Captures: [1] the newline or start, [2] the `#` run, [3] the heading text.
+ *
+ * @example
+ *   # Heading 1
+ *   ## Heading 2
+ *   ### Heading 3
+ */
 export const REGEXP_HEADER = /(^|\r\n|\r|\n)(#{1,6})\s+(.*)/gm;
 
+/**
+ * Matches level-2 headers only (`## …`).
+ *
+ * @example ## Heading 2
+ */
 export const REGEXP_H2 = /^##\s+(.*)$/gim;
 
+/**
+ * Matches level-3 headers only (`### …`).
+ *
+ * @example ### Heading 3
+ */
 export const REGEXP_H3 = /^###\s+(.*)$/gim;
diff --git a/src/tags/hr.ts b/src/tags/hr.ts
index 962cc7d..3c4420e 100644
--- a/src/tags/hr.ts
+++ b/src/tags/hr.ts
@@ -1,9 +1,6 @@
-import { NEWLINE } from '../utils';
-
 /**
- * Matches horizontal rule lines (5 or more dashes).
- * Platform-agnostic newline handling.
+ * Matches a horizontal rule: 5 or more dashes on their own line.
  *
- * Example: -----
+ * @example -----\n
  */
-export const REGEXP_HR = new RegExp(`${NEWLINE}-{5,}`, 'g');
+export const REGEXP_HR = /(?:\r\n|\r|\n)-{5,}/g;
diff --git a/src/tags/image.ts b/src/tags/image.ts
index ab8483c..bfc49d4 100644
--- a/src/tags/image.ts
+++ b/src/tags/image.ts
@@ -1,8 +1,7 @@
 /**
  * Matches markdown image syntax.
- * Captures: [1] alt text, [2] url
+ * Captures: [1] alt text, [2] url.
+ *
+ * @example ![alt text](https://example.com/image.png)
  */
-export const REGEXP_IMAGE = new RegExp(
-  '!\\[([^\\[]*?)\\]\\(([^)]+?)\\)',
-  'g'
-);
+export const REGEXP_IMAGE = /!\[([^\[]*?)\]\(([^)]+?)\)/g;
diff --git a/src/tags/index.js b/src/tags/index.js
deleted file mode 100644
index b5df3a9..0000000
--- a/src/tags/index.js
+++ /dev/null
@@ -1,13 +0,0 @@
-export { REGEXP_HEADER, REGEXP_H2, REGEXP_H3 } from './header.js';
-export { REGEXP_BLOCKQUOTE } from './blockquote.js';
-export { REGEXP_CODE } from './code.js';
-export { REGEXP_LINK } from './link.js';
-export { REGEXP_IMAGE } from './image.js';
-export { REGEXP_STRONG } from './strong.js';
-export { REGEXP_DEL } from './del.js';
-export { REGEXP_ITALIC } from './italic.js';
-export { REGEXP_HR } from './hr.js';
-export { REGEXP_BR } from './br.js';
-export { REGEXP_PARAGRAPH } from './paragraph.js';
-export { REGEXP_Q } from './quote.js';
-export { REGEXP_EMPTY_BLOCKQUOTE } from './empty-blockquote.js';
diff --git a/src/tags/italic.ts b/src/tags/italic.ts
index 7e806c3..e0dc7fe 100644
--- a/src/tags/italic.ts
+++ b/src/tags/italic.ts
@@ -1,2 +1,9 @@
-/** Italic: *text* or _text_, with surrounding whitespace or start/end. */
+/**
+ * Matches italic/emphasis using `*` or `_` delimiters.
+ * Requires surrounding whitespace (or start/end / `>` / `<`) so that
+ * mid-word underscores are less likely to match.
+ *
+ * @example  *italic*
+ * @example  _italic_
+ */
 export const REGEXP_ITALIC = /(^|[\s>])(\*|_)(.+?)\2([\s<]|$)/g;
diff --git a/src/tags/link.ts b/src/tags/link.ts
index 5ac888f..976ac1a 100644
--- a/src/tags/link.ts
+++ b/src/tags/link.ts
@@ -1,8 +1,7 @@
 /**
  * Matches markdown link syntax.
- * Captures: [1] link text, [2] url
+ * Captures: [1] link text, [2] url.
+ *
+ * @example [GitHub](https://github.com)
  */
-export const REGEXP_LINK = new RegExp(
-  '\\[([^\\[]*?)\\]\\(([^)]+?)\\)',
-  'g'
-);
+export const REGEXP_LINK = /\[([^\[]*?)\]\(([^)]+?)\)/g;
diff --git a/src/tags/paragraph.ts b/src/tags/paragraph.ts
index a7991de..cce1d15 100644
--- a/src/tags/paragraph.ts
+++ b/src/tags/paragraph.ts
@@ -1,7 +1,6 @@
-import { NEWLINE } from '../utils';
-
 /**
- * Matches paragraph content between newlines.
- * Platform-agnostic newline handling.
+ * Matches paragraph-like content sitting between two newlines.
+ *
+ * @example \\nhello world\\n
  */
-export const REGEXP_PARAGRAPH = new RegExp(`${NEWLINE}(.+?)${NEWLINE}`, 'g');
+export const REGEXP_PARAGRAPH = /(?:\r\n|\r|\n)(.+?)(?:\r\n|\r|\n)/g;
diff --git a/src/tags/quote.ts b/src/tags/quote.ts
index 030f9fd..8b3bf5e 100644
--- a/src/tags/quote.ts
+++ b/src/tags/quote.ts
@@ -1,7 +1,6 @@
 /**
- * Matches quoted text using the non-standard :"...": syntax
- * used by some email/newsletter pipelines.
+ * Matches the custom quote syntax used in some newsletter pipelines.
  *
- * Example: :"quoted text":
+ * @example :"quoted text":
  */
 export const REGEXP_Q = /:"(.*?)":/g;
diff --git a/src/tags/strong.ts b/src/tags/strong.ts
index 35d9c2c..301628f 100644
--- a/src/tags/strong.ts
+++ b/src/tags/strong.ts
@@ -1,2 +1,8 @@
-/** Bold: **text** or __text__. */
+/**
+ * Matches bold/strong text using `**` or `__` delimiters.
+ * Captures the inner text in group 1 (`**`) or group 2 (`__`).
+ *
+ * @example **bold**
+ * @example __bold__
+ */
 export const REGEXP_STRONG = /\*\*(.+?)\*\*|__(.+?)__/g;
diff --git a/src/utils.js b/src/utils.js
deleted file mode 100644
index 701fdc4..0000000
--- a/src/utils.js
+++ /dev/null
@@ -1,5 +0,0 @@
-import os from "os";
-
-const platform = os.platform();
-
-export const newLine = platform === "win32" ? "\r\n" : "\n";
diff --git a/src/utils.ts b/src/utils.ts
index e8e526c..03994d1 100644
--- a/src/utils.ts
+++ b/src/utils.ts
@@ -1,6 +1,6 @@
-/** Unix / Windows / old-Mac newline as a RegExp source fragment. */
-export const NEWLINE = String.raw`(?:\r\n|\r|\n)`;
-
-export function nl(pattern: string, flags = 'g'): RegExp {
-  return new RegExp(pattern.replace(/\n/g, NEWLINE), flags);
-}
+/**
+ * Flexible newline fragment: Unix (`\n`), Windows (`\r\n`), old Mac (`\r`).
+ * Prefer embedding `(?:\r\n|\r|\n)` directly in regex literals.
+ * This constant is kept for callers that compose patterns at runtime.
+ */
+export const NEWLINE = '(?:\r\n|\r|\n)';