What does this skill do, and when should you use it?
This is a Claude Skill that converts markdown content into properly formatted EPUB3 ebook files. It uses H1 headers as chapter boundaries and H2 as sections, automatically building a table of contents and navigation document, with metadata support including YAML frontmatter. Technical content gets extra care: code blocks use premium monospace fonts with styled backgrounds, and tables get styled headers and zebra striping. The resulting file can be downloaded and transferred to Kindle or any other reader. The README marks it version 1.0.0, production ready, and ships with a test suite.
Reads markdown pasted directly into a message or from a file path; markdown_processor.py parses heading hierarchy into chapters, extracts YAML frontmatter metadata, and converts markdown to HTML; epub_generator.py uses ebooklib to produce an EPUB3 file with TOC (NCX + NAV), metadata, and embedded CSS; outputs a downloadable .epub file.
- A researcher who wants to package markdown research notes into an ebook for offline reading on a tablet or Kindle.
- A developer exporting a long technical document while preserving code-block styling and table formatting.
- A user who wants to preserve an AI conversation summary as a portable reference document.
- A blogger assembling published articles into a chaptered book.
- A Python developer calling create_epub_from_markdown directly in scripts for batch conversion.
How do you install this skill?
- This is a static review; no code was executed. Verify actual EPUB output quality (list wrapping, nested lists, complex tables) before production use.
- markdown_processor.py is incomplete in the provided evidence; cross-check the full repository source and test_epub_skill.py results before installing.
- No Chinese-language documentation; language metadata defaults to 'en' — specify language manually for Chinese ebooks. Dependencies install via pip; assess reachability from mainland-China networks.
- README license wording conflicts with the MIT LICENSE file; treat LICENSE as authoritative.
- generate() raises on failure rather than returning False; handle exceptions when integrating.
- Shell / CLI
- Local filesystem
Python 3.8+pipebooklib 0.18.0markdown2 2.4.12
- Requires Python 3.8+ and pip. 2. Install dependencies: pip install -r markdown-to-epub/requirements.txt. 3. Verify with python test_epub_skill.py — markdown processing, EPUB generation, and edge cases should all PASS. Place the skill folder (markdown-to-epub/ containing SKILL.md) per the standard Claude skill location; the README does not document the exact install directory.
tmp="$(mktemp -d)"
git clone --depth 1 https://github.com/smerchek/claude-epub-skill.git "$tmp"
mkdir -p ~/.claude/skills
cp -R "$tmp/markdown-to-epub" ~/.claude/skills/
rm -rf "$tmp"Generated from the source repository and skill path; it copies only this skill's folder. If the author's install steps above differ, follow those first. To scope it to one project, replace ~/.claude/skills with that project's .claude/skills.
How do you use this skill?
Once installed, send your agent any of these to trigger it:
- Convert this markdown to EPUB: # My Book Title\n## Chapter 1\n...
- Convert the markdown file at research_notes.md to an EPUB file
- Summarize our conversation so far as markdown and convert it to an EPUB
As a skill, trigger it with natural language, e.g.: "Convert this markdown to EPUB: # My Book Title\n## Chapter 1\n..." or "Convert the markdown file at research_notes.md to an EPUB file"; you can also ask "Summarize our conversation so far as markdown and convert it to an EPUB". As a Python library: from markdown_to_epub.scripts.epub_generator import create_epub_from_markdown, passing markdown content plus output_path, title, and author. After generation, download the file and transfer it to Kindle, iPad, or another device.
What are this skill's strengths and limitations?
- EPUB3-compliant output compatible with Apple Books, Kindle apps, Google Play Books, Kobo, and standard readers.
- Automatic table of contents and chapter navigation from a simple H1/H2 convention.
- Enhanced code blocks and tables make it well suited to technical content.
- YAML frontmatter support for title, author, language, and date metadata.
- Ships with a test suite (test_epub_skill.py); README marks it production ready.
- No cover-page generation, .mobi output, or multi-document merging — all listed as future work.
- Chapter detection relies entirely on H1 headers; without them the TOC will be empty.
- License statement is inconsistent: README says 'same license as the Claude Cookbooks project' while the repo metadata says MIT — verify before adopting.
- Behavior on actual Kindle hardware is not validated in the source material, only the reading apps are listed.
- Depends on ebooklib and markdown2, requiring a local Python environment.
How does this skill compare with similar options?
Side by side with related skills; every score comes from the same FSRS standard.
| Skill | FS score | Stars | Last updated | License |
|---|---|---|---|---|
| Markdown to EPUB Skill this page | 52 · Use with care | ★ 167 | 1y ago | MIT |
| LLM Wiki Manager | 51 · Use with care | ★ 71 | 1mo ago | MIT |
| Nutrient Document Processing Skill | 69 · Recommended | ★ 18 | 1mo ago | — |
| PPT Master | 64 · Recommended | ★ 62k | 11d ago | MIT |
| huashu-md-html Publishing Pipeline | 57 · Use with care | ★ 915 | 12d ago | MIT |
How did FollowSkills review this skill?
Scripts are purely local: no network calls, no credential access, no covert exfiltration; HTML special chars are escaped and directory creation is a minimal side effect. Deductions: SKILL.md does not declare data flow or permission boundaries, no rollback guidance, publisher identity unverified, confirmation mechanisms undocumented.
Code is internally coherent, covering YAML frontmatter, code-block placeholder protection, and table parsing; errors are printed then re-raised with readable messages. Deductions: generate() docstring claims False on failure but actually raises (behavior/doc mismatch); markdown_processor.py is truncated in evidence so some paths cannot be checked; test_epub_skill.py is referenced but not provided in evidence; static review cannot confirm key-path reproduction.
Inputs (raw text/file path), H1 chapter rules, EPUB3 output, and trigger examples are clear. Deductions: no declared non-fit boundaries (e.g., no image embedding), no Chinese/multilingual note (language defaults to 'en'), mainland-China reachability of pip/ebooklib dependency installation is unevaluated.
Good layering (SKILL.md + REFERENCE.md), versioning, roadmap, contribution rules, and CI enforcing CHANGELOG entries plus tests before release. Deductions: README states license follows 'Claude Cookbooks project' which contradicts the MIT LICENSE; 'Maintained By: Skills Development Team' is vague ownership; markdown2 is listed in requirements but never used.
The pipeline (parse → chapters → EPUB3 with NCX/NAV and TOC) is complete and output should be directly usable, giving real marginal value over manual packaging. Deductions: static review cannot verify actual output quality; list rendering emits bare <li> without a wrapping <ul>/<ol>, and nested-list support is claimed but not visibly implemented, so output completeness is uncertain.
A release workflow installs dependencies, runs test_epub_skill.py, and enforces CHANGELOG, providing auditable primary material. Deductions: the test file itself is absent from the evidence, no third-party execution results or representative output samples exist, so reproduction coverage is thin under static review.
Open a dimension to read why it scored that way
Evidence confidence：Low — Mostly static review, author material or a limited demo; useful for discovery, not high-risk decisions.
See the full review method →