DeepRead Medical Records enables automated extraction of structured data from clinical documents while ensuring HIPAA-compliant PII redaction.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install deepread-medical
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install deepread-medical using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is DeepRead Medical Records?
DeepRead Medical Records is a specialized tool designed to bridge the gap between unstructured clinical documentation and digital healthcare systems. It leverages advanced OCR and medical-specific AI models to extract high-fidelity data from lab reports, prescriptions, and discharge summaries. By using this as part of your Openclaw Skills, you can automate the conversion of medical paperwork into structured JSON formats with 97%+ accuracy, significantly reducing manual data entry for EHR and research workflows.
Beyond simple extraction, this skill prioritizes security and compliance. It features context-aware PII redaction that distinguishes between institutional data (like provider names) and sensitive patient information (like names, SSNs, and dates of birth). This allows developers to generate HIPAA-compliant datasets suitable for training LLMs or sharing across non-clinical environments.
DeepRead Medical Records Use Cases
- Automating clinical data extraction from patient records into centralized EHR or data warehouses
- Converting complex laboratory reports into structured formats including test values and abnormal flags
- Digitizing handwritten or printed prescriptions into actionable medication data arrays
- De-identifying patient records for medical research datasets while preserving vital clinical insights
- Accelerating insurance claims processing by automatically pulling ICD-10 and CPT codes
How DeepRead Medical Records Works
- The document is submitted to the DeepRead API via a POST request with a defined extraction schema.
- DeepRead analyzes the document using medical-specific OCR to identify clinical entities and patient identifiers.
- Extracted data is mapped to the provided JSON schema, covering patient info, vitals, and diagnoses.
- Fields with lower confidence scores are marked with a human-in-the-loop (HIL) flag for manual verification.
- If requested, a context-aware redaction pass is performed to strip PII while maintaining the document's clinical utility.
- The final structured JSON and redacted files are retrieved through an asynchronous polling mechanism.
DeepRead Medical Records Setup
To integrate this capability into your Openclaw Skills environment, you must first obtain an API key from the DeepRead dashboard. Once acquired, configure your environment variables as follows:
export DEEPREAD_API_KEY="sk_live_your_key_here"
You can then utilize the provided Python or cURL examples to begin processing clinical PDFs or images immediately.
DeepRead Medical Records Data Schema & Taxonomy
The skill organizes clinical data into a comprehensive hierarchy, ensuring metadata taxonomy is preserved for medical records. The schema includes:
| Category | Fields Included |
|---|---|
| Patient Identity | name, date_of_birth, mrn (medical record number) |
| Clinical Vitals | blood_pressure, heart_rate, temperature, weight, height |
| Diagnoses | ICD-10 codes and detailed clinical descriptions |
| Medications | drug name, dosage, frequency, and duration |
| Labs | test_name, value, reference_range, and abnormal flags |
| Redaction | 14+ PII types including SSNs, emails, and addresses |
DeepRead Medical Records Advanced Features
- Context-aware PII detection that distinguishes between patient data and institutional headers
- Human-in-the-loop (HIL) flagging to ensure 100% accuracy on critical medical fields
- Bring Your Own Key (BYOK) support for OpenAI, Google, or OpenRouter to reduce processing costs
- Custom blueprint training for recurring form types to improve extraction accuracy by up to 30%
- Seamless integration with other Openclaw Skills for end-to-end medical automation workflows
SKILL.md
Loading
Related Openclaw Skills
AntV G2 v5 Chart Generator
A professional-grade toolkit for generating accurate, high-performance data visualizations using the AntV G2 v5 library and Spec Mode syntax.
Mopidy Party Mode
A collaborative music management skill that allows group chat members to contribute to a Mopidy queue without disrupting playback.
Zhihuiya Patent Description Data
A professional tool for retrieving technical patent specifications and detailed technical descriptions directly from the Zhihuiya database.
Zhihuiya Patent Description (Translated)
A professional tool for fetching and translating full patent specifications into Chinese, English, or Japanese using Zhihuiya data.
Changan Information Assistant
A specialized AI skill for retrieving real-time product data, financial reports, and official corporate news for Changan Automobile.
Mopidy
A sophisticated interface for AI agents to control Mopidy music systems via JSON-RPC for seamless playback and library management.
Featured*
