An interactive PDF reader that enables AI agents to navigate, search, and process documents through block-level commands and natural language.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install book-walker
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install book-walker using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is book-walker?
The book-walker skill is an interactive PDF reading engine designed specifically for deep document interaction. It allows AI agents and users to navigate through PDFs sentence by sentence or block by block, providing a more granular and controlled reading experience than standard viewers. As a core component of Openclaw Skills, it provides the bridge between raw document data and intelligent processing, allowing agents to maintain state, remember progress, and handle complex layouts effectively.
Built for high-precision workflows, this skill supports two distinct extraction modes: a fast text-based mode for standard digital documents and a robust OCR mode for scanned images. By breaking documents into logical segments, it ensures that AI models can process content without being overwhelmed by excessive context, making it an ideal choice for technical documentation, academic research, and long-form report analysis.
book-walker Use Cases
- Deeply reading academic papers or technical manuals sentence by sentence to ensure full comprehension.
- Navigating large PDF libraries within a workspace using automated indexing and keyword search.
- Processing scanned or legacy documents that require OCR for accurate text extraction.
- Managing multi-session research tasks by setting bookmarks and tracking reading progress across different files.
- Integrating LLM-based translation or summarization workflows directly into the reading process using custom templates.
How book-walker Works
- The user initiates the skill by either listing PDFs in the workspace or opening a specific file path.
- The engine parses the PDF into logical blocks (paragraphs) and creates a local index and cache for fast retrieval.
- Navigation commands like Next, Previous, or Jump allow the user to move through the document while the skill maintains a persistent state.
- When using specific templates, the skill generates a structured payload that includes the original text and a prompt, which the AI agent then processes via an LLM.
- Progress is saved automatically to a local cache directory based on the PDF's unique hash, allowing users to resume reading at any time.
book-walker Setup
To integrate this into your Openclaw Skills environment, ensure Python 3 is available and install the necessary dependencies:
cd ~/.openclaw/workspace-e/skills/book-walker
python3 -m venv .venv
.venv/bin/pip install pdfplumber pypdfium2
For OCR functionality, Tesseract must be installed on your system:
# macOS
brew install tesseract
# Ubuntu
sudo apt install tesseract-ocr
book-walker Data Schema & Taxonomy
The book-walker skill organizes its data within ~/.cache/pdf-reader/{hash}/ to ensure high performance and persistence. The schema includes:
| File | Purpose |
|---|---|
index.json |
Contains total page counts and byte offsets for navigation. |
p[n].json |
Stores the extracted text blocks for a specific page 'n'. |
state.json |
Tracks the current page, block ID, and overall reading percentage. |
bookmarks.json |
Stores user-defined bookmarks with associated notes and timestamps. |
book-walker Advanced Features
- Hybrid Extraction Engine: Toggle between text-based extraction for speed and OCR-based extraction for scanned content.
- LLM Processing Templates: Define custom templates like 'Translate' or 'Explain' which the AI agent applies to the active text block.
- Smart Text Enhancement: Features an intelligent post-processing module that fixes common PDF issues like missing spaces in LaTeX-generated files.
- Multi-Granular Navigation: Support for jumping directly to specific pages, blocks, or lines using natural language triggers.
- Workspace Indexing: Automatically scans and indexes all PDF files within the environment for quick access via short identifiers.
SKILL.md
Loading
Related Openclaw Skills
MarkItDown Document Converter
A high-performance document conversion tool that leverages Microsoft's MarkItDown library to turn various file formats into structured Markdown.
Gemini Image Gen
A high-performance AI image generation tool utilizing Google Gemini models for text-to-image and reference-guided design.
Self-Criticism
A strategic error-correction skill that inserts breakpoint self-critique at risky inflection points to prevent downstream mistake propagation.
Careers Page Generator
A specialized tool for AI agents to design, audit, and optimize high-converting careers pages and employer branding content.
Seedance Promptor
An expert AI assistant for generating technical, multimodal storyboard prompts specifically optimized for the Seedance 2.0 video engine.
Tavily Search
A specialized web search tool for AI agents that delivers concise, context-rich results and summaries through the Tavily API.
Featured*
