WiseDiag MedOCR
ReviewAudited by ClawScan on May 10, 2026.
Overview
WiseDiag MedOCR appears to do what it says—OCR one selected file through WiseDiag’s cloud API—but users should notice the cloud upload, API key use, and unpinned Python dependencies.
Before installing, make sure you trust WiseDiag with the documents you plan to OCR and with the API key. Do not use this for sensitive files unless you accept remote processing; install dependencies in an isolated environment if possible.
Findings (3)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
Any PDF or image processed with this skill may be transmitted to WiseDiag and processed remotely.
The code uploads the selected file to WiseDiag’s cloud OCR endpoint. This is central to the skill and is disclosed in the documentation, but it means document contents leave the local machine.
DEFAULT_SERVICE_URL = "https://openapi.wisediag.com" ... resp = requests.post(endpoint, files=multipart, params=params, headers=headers, timeout=REQUEST_TIMEOUT)
Use it only for documents you are comfortable sending to WiseDiag; use an offline OCR tool for confidential, medical, financial, credential, or personal documents.
The API key authorizes OCR requests under the user’s WiseDiag account and may consume quota or expose account access if mishandled.
The skill requires a WiseDiag API key. This is expected for the provider API, and the artifacts do not show hardcoding, logging, or unrelated use of the key.
env_vars:\n - WISEDIAG_API_KEY\ncredentials:\n required: true
Set the API key only in trusted environments, avoid committing it to files or sharing shell profiles, and rotate it if it may have been exposed.
A later or different dependency version could behave differently from what was reviewed here.
The install dependencies are normal for this Python OCR wrapper, but they are not pinned to exact versions, so future installs may fetch versions not represented in the reviewed artifacts.
requests>=2.28.0\npypdf>=3.0.0
Install in a virtual environment and consider pinning or locking dependency versions if you need reproducible or high-assurance installs.