System Dependencies

• - uv must already be installed because this skill is executed with uv run, and uv installs the Python dependencies declared in src/main.py.
• INLINECODE4 is needed for tts because the speech output is normalized and written as an .mp3 file through ffmpeg.
• INLINECODE8 is needed for ocr because it performs the actual optical character recognition on scanned page images.
• INLINECODE10 is also needed for ocr because it extracts page images from PDFs before those images are passed to tesseract; pdfimages comes from poppler.
• INLINECODE15 is optional for convert because it can convert between many document formats when text-based conversion is possible.
• INLINECODE17 is an optional alternative to pandoc for convert because it can handle document conversions that pandoc may not support well.

File Access And Network Behavior

• - This skill operates on the file paths provided by the caller. It can read from and write to any host path the caller supplies; it is not limited to the OpenClaw workspace.
• The /root/.openclaw/workspace/... paths in the command examples show where the skill entrypoint lives. They do not restrict which files the skill can access.
• The tts command uses edge-tts, which sends the input text to an external text-to-speech service over the network to generate audio.
• Do not use tts with sensitive or private text unless you are comfortable sending that text off-host.
• All other commands run locally on the host, subject to the optional local binaries documented below.

Skill: PDF Toolkit

When to use

• - User wants to extract text, tables, or images from a PDF.
• User wants to get metadata or page count from a PDF.
• User wants to merge, split, or rotate a PDF.
• User wants to create a new PDF from plain text or Markdown.
• User wants to read or write a DOCX file.
• User wants to OCR a scanned PDF (requires tesseract on host).
• User wants to convert text or a document to an MP3 audio file (requires ffmpeg on host).
• User wants to convert between document formats (requires pandoc or libreoffice on host).
• User wants to check which optional system tools are available.

When NOT to use

• - User wants to view or render a PDF visually — use a PDF viewer.
• User wants to fill in PDF form fields — this skill does not support AcroForms.
• User wants to edit an existing PDF's text in-place — use a dedicated PDF editor.

Commands

Check available tools

Get PDF metadata and page count

Extract text from a PDF

Extract tables from a PDF

Extract images from a PDF

Merge PDFs

Split a PDF

Rotate pages in a PDF

Create a PDF from text

Read a DOCX file

Write a DOCX file

OCR a scanned PDF (requires tesseract)

Convert text or document to speech (requires ffmpeg)

Convert document formats (requires pandoc or libreoffice)

Examples

CODEBLOCK14

Chat Delivery

• - When this skill is used in a chat interface that supports file attachments, such as Telegram, any generated output file should be sent back to the user as an attachment after successful creation or conversion.
• This applies to commands that create files, including create-pdf, write-docx, extract-images, merge, split, rotate, tts, and convert.
• If a temporary output file is created in the Claw runtime temporary folder for delivery, delete that temporary file immediately after the file has been sent successfully to the user.
• Do not delete files that were written to a user-requested destination outside the Claw temporary folder.
• If the chat environment cannot send file attachments, report the output path clearly instead of claiming the file was delivered.

Output

• - Plain text with labeled sections separated by blank lines.
• Errors are prefixed with Error:.
• The doctor command shows a table of available and missing tools.

Notes

• - uv run reads the inline # /// script dependency block in main.py and auto-installs Python packages in an isolated environment — no pip install or venv setup needed.
• Core features (info, extract-text, extract-tables, merge, split, rotate, create-pdf, read-docx, write-docx) work with uv alone — no system binaries required.
• OCR requires tesseract installed on the host (brew install tesseract / apt install tesseract-ocr). Also needs pdfimages from poppler (brew install poppler).
• TTS requires ffmpeg installed on the host (brew install ffmpeg / apt install ffmpeg).
• Document conversion requires pandoc or libreoffice on the host.
• Run doctor first if you are unsure which features are available.