Clinical Data Extractor

What This Skill Does

The Clinical Data Extractor is a specialized OpenClaw agent skill designed for medical researchers, pharmaceutical analysts, and healthcare professionals. It automates the tedious process of scouring pharmaceutical conference websites (such as ASCO, ESMO, and EHA) and PDF documents to identify and structure data from innovative drug clinical trials.

By leveraging the built-in browser and file system tools, this skill parses complex medical documents and converts them into clean, standardized Markdown files. It captures essential metadata including the drug name, manufacturer, indication, clinical phase, trial name, conference details, and comprehensive efficacy and safety data presented in clear tables. The skill applies specific naming conventions to ensure all files are organized systematically within your workspace.

Installation

To integrate this skill into your OpenClaw environment, execute the following command in your terminal: clawhub install openclaw/skills/skills/abinww/clinical-data-extractor Ensure your host machine has Google Chrome installed, as the skill utilizes the browser dependency for dynamic content extraction. If you plan to process PDF files, ensure poppler-utils is installed on your system to provide robust text extraction capabilities.

Use Cases

This skill is ideal for:

• Summarizing drug efficacy data from large conference posters or slide decks.
• Batch-processing clinical trial results from pharmaceutical company websites.
• Converting unstructured, scattered trial data into organized Markdown reports for internal knowledge bases.
• Standardizing reporting formats across different therapeutic areas and drug trials.

Example Prompts

1. "Extract the clinical trial data from this link: https://www.asco.org/clinical-trial-abstract-12345 and save it as a markdown file."
2. "Please process the PDF document at ~/Downloads/study_report.pdf. Focus on extracting the safety tables and OS/PFS outcomes."
3. "Go to the ESMO conference page for the Lung Cancer session, identify the Phase 3 trial for Drug X, and create a structured summary in the workspace."

Tips & Limitations

For optimal results, ensure the target website is not protected by heavy anti-bot measures. If the skill fails to capture data from a secure or interactive platform like a WeChat article, manual screenshots are recommended. Always check the generated markdown file to ensure abbreviations were parsed according to your configured list. The skill performs best when the source material is high-quality text-based content rather than low-resolution images.