WiseOCR

PDF & Image OCR — Convert a single PDF or image to Markdown via WiseDiag cloud API. Supports table recognition, multi-column layouts, and high-accuracy text...

Security Scan

The skill's code, instructions, and requested secret (WISEDIAG_API_KEY) are coherent with its stated purpose (upload a single PDF/image to WiseDiag for OCR); nothing indicates behavior beyond that.

Purpose & Capability

The name/description claim cloud OCR via WiseDiag and the included script, README, and SKILL.md all implement exactly that (uploads a file to https://openapi.wisediag.com/v1/ocr/pdf and returns markdown). One inconsistency: registry metadata listed 'Required env vars: none' and 'Primary credential: none', but the SKILL.md, README, and script clearly require WISEDIAG_API_KEY.

Instruction Scope

Runtime instructions are limited and explicit: set WISEDIAG_API_KEY, run scripts/wiseocr.py on a single file, and the tool uploads the file to WiseDiag and saves returned Markdown locally. The SKILL.md explicitly warns about privacy and forbids calling other endpoints directly. The script only reads the input file, the WISEDIAG_API_KEY env var, and writes output to ~/.openclaw/workspace/WiseOCR — no out-of-scope file/system access is present.

Install Mechanism

No automated install spec is provided (instruction-only install via pip). requirements.txt references standard packages (requests, pypdf). No downloads from untrusted URLs, no archives extracted, and no custom binary locations — proportional to the task.

Credentials

The script requires a single API credential (WISEDIAG_API_KEY), which is appropriate for a cloud OCR service. However, the registry metadata incorrectly reports no required env vars/credentials; that mismatch should be corrected before trusting the registry listing. No other secrets or unrelated credentials are requested.

Persistence & Privilege

The skill does not request permanent 'always' inclusion, does not modify other skills or system-wide settings, and only writes output into its own workspace directory (~/.openclaw/workspace/WiseOCR). Autonomous invocation defaults are unchanged.

Assessment

This skill appears to do exactly what it says: it uploads a single PDF/image to WiseDiag's API and saves returned Markdown locally. Before installing: (1) Be aware that files are uploaded to a third-party server — do not use with sensitive or confidential documents unless you trust WiseDiag's policies. (2) The script requires an API key set in WISEDIAG_API_KEY; verify how you store that secret (environment variable vs. long-lived shell files). (3) Note the registry metadata incorrectly omits the required env var — treat the SKILL.md/README as authoritative and prefer inspecting the script (scripts/wiseocr.py) yourself if you have doubts. (4) If you need offline processing for sensitive data, use a local OCR tool instead.

⚠️ Privacy Warning

IMPORTANT - READ BEFORE INSTALLING:

This skill uploads your file to WiseDiag's cloud servers for OCR processing.

Do NOT use with sensitive or confidential documents unless:

• You trust WiseDiag's data handling policies
• You accept that file contents will be transmitted and processed remotely

For sensitive documents, use offline/local OCR tools instead.

---

WiseOCR Skill (powered by WiseDiag)

A high-accuracy OCR tool that converts a single PDF or image file into Markdown format. After processing, the Markdown result is automatically saved to disk — no additional saving is needed.

Supported formats: PDF, jpg, jpeg, png, webp, gif, bmp, tiff.

Installation

pip install -r requirements.txt

🔑 API Key Setup (Required)

Get your API key: 👉 https://console.wisediag.com/apiKeyManage

The API key MUST be set as an environment variable. The script reads it automatically.

export WISEDIAG_API_KEY=your_api_key

How to Process a File (Step-by-Step)

NEVER call any API or HTTP endpoint directly. ONLY use the script below.

Step 1: Set the API key (if not already set):

export WISEDIAG_API_KEY=your_api_key

Step 2: Run the script with the input file:

cd scripts

# Single PDF
python3 wiseocr.py -i "/path/to/input_filename.pdf"

# Single image
python3 wiseocr.py -i "/path/to/scan.png"

IMPORTANT: If the input file has been copied or renamed (e.g. to a temp path), always pass -n with the original filename (without extension) so the output file is named correctly:

python3 wiseocr.py -i "/tmp/ocr_input.pdf" -n "my_report"
# Output saved to: ~/.openclaw/workspace/WiseOCR/my_report.md

The Markdown result is saved to ~/.openclaw/workspace/WiseOCR/{name}.md automatically. No additional saving is needed.

Arguments

Flag	Description
-i, --input	Input file: PDF or image — single file path (required)
-n, --name	Output filename stem (recommended when input file is renamed/copied)
-o, --output	Output directory (default: ~/.openclaw/workspace/WiseOCR)
--dpi	PDF rendering DPI, 72-600 (default: 200)

Data Privacy

What happens to your files:

1. Files are uploaded to WiseDiag's OCR API
2. Files are processed on WiseDiag servers
3. Processing results are returned to you
4. Files are not permanently stored on WiseDiag servers

For sensitive documents, use offline/local OCR tools instead.

License

MIT