book-to-skill

An agent skill by virgiliojr94, from virgiliojr94/book-to-skill. Tags: anthropic, automation, cli, developer-tools, documents.

What it does

Converts books and documents (PDF, EPUB, DOCX, HTML, Markdown, plain text, RTF, MOBI/AZW with Calibre) into structured agent skills, extracting frameworks, mental models, principles, techniques, and anti-patterns. Use when the user wants to study a document through GitHub Copilot CLI, Amp, Claude Code, Hermes Agent, or OpenClaw, apply an author's frameworks while working, or build a reusable knowledge base from a file.

Install

With the skills CLI, which installs into Claude Code, Codex, Cursor and other agents:

npx skills add virgiliojr94/book-to-skill --skill book-to-skill

Or copy the skill folder into Claude Code's skills directory by hand (~/.claude/skills for every project, or .claude/skills inside one):

git clone --depth 1 https://github.com/virgiliojr94/book-to-skill
cp -r book-to-skill ~/.claude/skills/book-to-skill

Safety box score

Grade D, elevated risk: 51 of 100 risk points, rated 2026-10-15. The grade rates what the skill and its scripts can reach on the machine of whoever installs it, across eight categories from shell execution to secrets access. How the score works.

This skill converts local books/documents into agent skills, writing chapter files plus indexes into a chosen skills root and optionally publishing the result to GitHub. The main risks are subprocess-heavy extraction, optional package/network activity, and broad filesystem writes including recursive cleanup of the per-run workdir.

Category levels, from 0 (none) to 3 (high)
Category	Level	Why
Code & shell execution	2 (moderate)	It runs its own Python entrypoints and external converters, e.g. `"$PYTHON_BIN" "$SCRIPT_PATH" ... --install-missing ask` and `subprocess.run(["ebook-convert", ...])` in `book_to_skill/parsers/calibre.py`.
Network access	2 (moderate)	The workflow can contact package registries and GitHub, including `pip install "book-to-skill[pdf,epub,docx] @ git+https://github.com/virgiliojr94/book-to-skill.git"`, `gh repo create ... --push`, and `npx skills add ...`.
Filesystem reach	2 (moderate)	It writes generated skills into user or project skill roots such as `~/.agents/skills/<skill_name>` and explicitly removes the extraction workdir with `rm -rf "$WORKDIR"` / `shutil.rmtree(...)`.
Secrets & credentials	1 (low)	It does not read keychains or `.env` files, but optional publishing is done through the user's authenticated GitHub CLI session (`gh auth status` / `gh repo create --push`).
Privilege & persistence	1 (low)	No sudo or system service install is present, but it may create symlinks in user directories and ask to install missing user-space tools like `gh` or Python packages.
Instruction hijack surface	1 (low)	`SKILL.md` is heavily instruction-shaped and uses HTML comments and host-specific guidance, but I did not see payloads that try to override safeguards, hide actions, or follow remote instructions.
Irreversible actions	2 (moderate)	The skill can create/update skills and, if the user opts in, publish and push a repository (`gh repo create ... --push`) and offer `npx skills add` installation.
Transparency	2 (moderate)	Several substantial runtime files were not readable in this sample (`book_to_skill/utils.py`, `book_to_skill/parsers/pdf.py`, `book_to_skill/sanitize.py`, etc.), so the full extraction behavior could not be audited.

Before installing

Read `SKILL.md` Step 11 if you might publish; it can create and push a GitHub repo through your logged-in `gh` account.
Confirm the destination root (`~/.agents/skills`, `~/.claude/skills`, or project-local) before first run; it writes there and only deletes its own temp workdir.
Run a scratch conversion with `BOOK_SKILL_WORKDIR` set to a temp path first if you want to see whether optional extractors or external converters are invoked.

Rated 2026-10-15 by an automated review of 70 files from the skill's repository.

Source

Repository: virgiliojr94/book-to-skill (all skills from this repository)
Path: SKILL.md
Branch: master
Updated: 2026-09-19

filesystem-context — This skill should be used when agent work needs file-backed context: durable scratchpads, tool-output offloading, just-in-time discovery.
cli-for-agents — Designs or reviews CLIs so coding agents can run them reliably: non-interactive flags, layered --help with examples, stdin/pipelines, fast actionable errors.
modern-python — Configures Python projects with modern tooling (uv, ruff, ty). Use when creating projects, writing standalone scripts, or migrating from pip/Poetry/mypy/black.
changelog-video — Turn a weekly changelog .md into a finished branded changelog video (square 1080, ~45-60s, Annie VO, animated brand background, mock-UI visualizations.
cli-developer — Use when building CLI tools, implementing argument parsing, or adding interactive prompts.
context-compression — This skill should be used when long-running agent sessions need context compression, structured summarization, compaction, token-per-task optimization.