メニュー
Convert and extract text from PDFs, DOCX, images (OCR), and other document formats using the gateway's built-in document processing stack.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Manage the Trinity AGI auth service and RBAC system -- database schema, role hierarchy, permission management, JWT flow, audit logging, and API endpoints.
Sync files between the host machine and the running Trinity Docker containers — config, agent files, skills, and extensions.
Develop the Trinity AGI Flutter Web Shell -- understand the architecture, A2UI renderer, WebSocket models, dual-client pattern, design tokens, and build workflow.
Operate the Trinity AGI Kubernetes infrastructure -- Helm charts, multi-tenant per-user OpenClaw pods, Vault Agent Injector secrets, minikube dev, and production deployment.
Operate the OpenClaw Gateway that powers Trinity AGI — configure providers, manage sessions, use the CLI, and understand the WebSocket protocol.
Manage the Trinity AGI Docker Compose stack — start, stop, logs, rebuild frontend, and full deploy.
| name | doc-convert |
| description | Convert and extract text from PDFs, DOCX, images (OCR), and other document formats using the gateway's built-in document processing stack. |
| homepage | https://github.com/trinityagi/trinity |
| metadata | {"openclaw":{"emoji":"📑","requires":{"bins":["pdftotext","pandoc","tesseract"]}}} |
Extract text from PDFs, DOCX files, images (via OCR), and other document formats. All tools are pre-installed in the OpenClaw gateway container.
| Tool | Binary | Purpose |
|---|---|---|
| pdftotext | pdftotext (poppler-utils) | Extract text from PDF files |
| pdfinfo | pdfinfo (poppler-utils) | Get PDF metadata (page count, size, etc.) |
| pdfimages | pdfimages (poppler-utils) | Extract images from PDFs |
| pandoc | pandoc | Convert between formats (DOCX->text, HTML->md, etc.) |
| tesseract | tesseract | OCR - extract text from images |
| python3 | python3 | pdfplumber, PyPDF2, python-docx, Pillow, pytesseract |
| libreoffice | libreoffice | Convert DOCX/XLSX/PPTX to PDF |
| imagemagick | convert / identify | Image manipulation and format conversion |
pdftotext input.pdf - # stdout
pdftotext -layout input.pdf output.txt # preserve layout
pdftotext -f 1 -l 5 input.pdf - # pages 1-5 only
pdfinfo input.pdf
pandoc -f docx -t plain input.docx # stdout
pandoc -f docx -t markdown input.docx # as markdown
libreoffice --headless --convert-to pdf input.docx
tesseract image.png stdout # extract text from image
tesseract image.png output -l eng pdf # OCR to searchable PDF
tesseract image.png stdout -l chi_sim # Chinese OCR
pandoc -f html -t markdown input.html
# PDF with tables
import pdfplumber
with pdfplumber.open("file.pdf") as pdf:
for page in pdf.pages:
print(page.extract_text())
for table in page.extract_tables():
print(table)
# DOCX
from docx import Document
doc = Document("file.docx")
for para in doc.paragraphs:
print(para.text)
# Image OCR via Python
from PIL import Image
import pytesseract
text = pytesseract.image_to_string(Image.open("image.png"))
print(text)
| Format | Read | Convert To |
|---|---|---|
| pdftotext, pdfplumber, PyPDF2 | text, markdown, images | |
| DOCX | pandoc, python-docx | text, markdown, PDF |
| HTML | pandoc | text, markdown, PDF |
| Images (PNG/JPG/TIFF/BMP) | tesseract (OCR) | text, searchable PDF |
| XLSX/PPTX | libreoffice | PDF, then text |
| RTF | pandoc | text, markdown |
| EPUB | pandoc | text, markdown |
| ODT | pandoc, libreoffice | text, PDF |
tesseract or pdfplumber with image extraction.pdftotext is fastest for text-based PDFs. Use pdfplumber when you need table extraction.pandoc handles format conversion between most document types.pdftotext -f 1 -l 10) to avoid memory issues.