← Back to Skills Marketplace
944
Downloads
5
Stars
1
Active Installs
28
Versions
Install in OpenClaw
/install wiseocr
Description
PDF & Image OCR — Convert a single PDF or image to Markdown via WiseDiag cloud API, with high-accuracy text extraction, table recognition, and multi-column l...
Usage Guidance
This skill appears to be what it says: a client that uploads a single PDF/image to WiseDiag and saves the returned Markdown locally. Before installing, consider the following: 1) The SKILL.md/README and script require WISEDIAG_API_KEY — set this only if you trust the WiseDiag service; the top-level registry metadata incorrectly omitted this requirement, so expect to provide the env var. 2) Do NOT upload sensitive/confidential documents — the skill uploads files to https://openapi.wisediag.com for processing. 3) Inspect the small script yourself (it's included) if you want to verify there are no extra network endpoints or telemetry; it uses only requests.post to the declared endpoint. 4) If you need offline processing for privacy, use a local OCR tool instead. 5) If you plan automated use, consider limiting the API key permissions on the WiseDiag console (if supported) and rotate keys regularly.
Capability Analysis
Type: OpenClaw Skill
Name: wiseocr
Version: 1.0.28
The WiseOCR skill is a legitimate tool for converting PDFs and images to Markdown using the WiseDiag cloud API. The core script (scripts/wiseocr.py) performs standard file handling and HTTP requests to the documented endpoint (openapi.wisediag.com) without any signs of obfuscation, persistence, or unauthorized data access. Both SKILL.md and README.md include prominent privacy warnings regarding the cloud-based nature of the service and advise against processing sensitive personal information.
Capability Assessment
Purpose & Capability
Name, description, README, SKILL.md and the script all consistently implement a single-file PDF/image OCR that uploads to WiseDiag and returns Markdown. The code performs file validation, page counting, upload, and writing output to ~/.openclaw/workspace/WiseOCR — all proportional to an OCR client.
Instruction Scope
SKILL.md instructs the user to set WISEDIAG_API_KEY and to run the provided script; the script only reads the provided input file, uses pypdf to count pages (for PDFs), uploads the file to the declared API endpoint, and writes back a markdown file. There are no instructions to read unrelated system files or exfiltrate other secrets. The README and SKILL.md explicitly warn about privacy and uploading sensitive documents.
Install Mechanism
This is instruction-only with a small Python script and requirements.txt; there is no install spec that downloads arbitrary archives or runs code from unknown URLs. The only external dependency is requests and pypdf via pip — expected for a Python CLI.
Credentials
The script and docs require a single environment variable WISEDIAG_API_KEY (reasonable for a cloud OCR service). However, the registry metadata at the top of the provided summary lists 'Required env vars: none' and 'Primary credential: none', which contradicts SKILL.md/README and the script (which exits if WISEDIAG_API_KEY is unset). This appears to be a packaging/metadata inconsistency and should be corrected but does not change the proportionality of the credential.
Persistence & Privilege
The skill is not marked always:true and does not modify other skills or system-wide agent settings. It writes output to a workspace directory under the user's home (~/.openclaw/workspace/WiseOCR) which is normal for a CLI tool. Autonomous invocation is allowed by default (expected) but that is not combined with broad unexpected privileges.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install wiseocr - After installation, invoke the skill by name or use
/wiseocr - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.28
Version 1.0.28 of WiseOCR
- No file changes detected in this release.
- Functionality and documentation remain unchanged.
v1.0.27
- Updated the privacy warning section: Medical & health information is no longer explicitly listed as restricted data.
- Documentation now omits references to medical records, diagnoses, prescriptions, and health insurance information in privacy and sensitive information warnings.
- All guidance and usage instructions remain unchanged.
v1.0.26
wiseocr 1.0.26
- No code or documentation changes detected in this version.
- Functionality and usage instructions remain the same.
v1.0.25
- Updated privacy warning: medical/health information examples removed from sensitive data list.
- Clarified sensitive categories in warnings to match latest recommendations.
- No functional or API changes; documentation update only.
v1.0.24
wiseocr 1.0.24
- No code or documentation changes detected in this release.
- Functionality and privacy warnings remain unchanged.
v1.0.23
- Removed references to "Medical & health information" from the sensitive information warnings.
- Updated privacy tables and lists to exclude medical-related examples.
- No functional or code changes; documentation update only.
v1.0.22
- Updated API key management link in documentation to use the correct WiseDiag Console address.
- No code or functional changes; documentation clarity improvement only.
v1.0.21
- Skill name updated from "wisediag-medocr" to "wiseocr" for consistency.
- Description revised and slightly shortened for clarity; core OCR features remain unchanged.
- No code changes or feature updates; documentation only.
- All usage instructions, privacy warnings, and installation steps remain the same.
v1.0.20
No user-visible changes in this release.
- No file changes detected since the previous version.
v1.0.19
- Added a detailed warning section specifying categories of personal sensitive information not to be uploaded (e.g., identity, financial, medical, biometric, credentials, minors' info).
- Reinforced privacy guidance to recommend local/offline OCR for documents with any sensitive or regulated data.
- Clarified existing privacy instructions with explicit examples and a warning table for better user understanding.
v1.0.18
No user-facing changes in this release.
- No file changes or updates detected between versions.
- Functionality and documentation remain the same.
v1.0.17
- Added support for image OCR (jpg, jpeg, png, webp, gif, bmp, tiff) alongside PDF.
- Clarified usage: now processes a single PDF or image file per request.
- Updated instructions for file types and usage examples.
- Refined privacy notice and installation/setup instructions.
- Minor rewording and structure improvements for clarity.
v1.0.16
No user-facing changes detected in this version.
- No file or documentation updates present.
- Functionality and usage remain the same as the previous release.
v1.0.15
- Improved privacy documentation: clarified data privacy and file handling steps in SKILL.md.
- Updated API key setup instructions and registration link for easier access.
- Streamlined usage instructions, emphasizing use of the provided script for processing.
- Shortened and clarified the skill description to clearly state cloud upload and required credentials.
v1.0.14
- No changes detected in this version.
- Functionality, documentation, and configuration remain unchanged.
v1.0.13
No changes were detected in this version.
- No file changes or updates were made for version 1.0.13.
v1.0.12
- Added a prominent privacy warning about file uploads to WiseDiag's cloud servers.
- Updated description to highlight support for table recognition, multi-column layouts, and improved text extraction.
- Clarified that files should not be used if containing sensitive or confidential information unless the user trusts WiseDiag.
- No code changes; documentation and safety guidance improvements only.
v1.0.11
- Clarified that PDF OCR is performed via WiseDiag's cloud API, with files uploaded for processing.
- Added a new Data Privacy & Cost section explaining API usage fees and privacy implications.
- Advised users to avoid uploading confidential documents unless they trust WiseDiag, and to consider offline alternatives for sensitive files.
- Updated description to mention potential API costs and file upload requirement.
v1.0.10
Version 1.0.10 Changelog
- Updated name and author details for consistency with upstream project.
- Simplified and clarified usage instructions and API key setup.
- Modified default output directory to `~/.openclaw/workspace/WiseOCR`.
- Streamlined privacy notice and removed duplicate warnings.
- Improved documentation layout and reduced repetition.
v1.0.9
- Updated SKILL.md to remove Chinese examples and keywords, using only English filenames and terms for usage instructions and examples.
- Generalized usage documentation (e.g., example filenames like "medical_report" instead of "体检报告").
- No changes were made to the code or functionality.
Metadata
Frequently Asked Questions
What is WiseOCR?
PDF & Image OCR — Convert a single PDF or image to Markdown via WiseDiag cloud API, with high-accuracy text extraction, table recognition, and multi-column l... It is an AI Agent Skill for Claude Code / OpenClaw, with 944 downloads so far.
How do I install WiseOCR?
Run "/install wiseocr" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is WiseOCR free?
Yes, WiseOCR is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does WiseOCR support?
WiseOCR is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created WiseOCR?
It is built and maintained by wisediag (@wisediag); the current version is v1.0.28.
More Skills