Deep Health Report Interpreter

From Raw Values to Actionable Guidance

Health reports usually contain many lab values across blood counts, liver function, kidney function, lipids, and glucose. When a user sees flags such as high, low, or arrow symbols, it is easy to focus on one abnormal item and miss the clinical context and priority. This skill targets non-clinical users and addresses the problem of understanding which abnormal values matter, whether follow-up is needed, and what lifestyle changes are reasonable. It explicitly does not provide diagnosis, treatment advice, or medication recommendations, and it turns a report into structured, readable guidance.

How It Works

The skill supports three input modes: uploading report screenshots or photos, typing indicator names and values, or a hybrid flow with key pages photographed and remaining items typed. The workflow has three main steps:
- Extract and organize: identify indicator names, test results, and reference ranges, then organize them into a four-column table for review.
- Risk grading: use common reference ranges and thresholds to classify abnormal or borderline values into yellow or red risk levels, and describe follow-up or consultation directions.
- Generate interpretation: focus on abnormal and borderline items, mark normal items as unremarkable, and provide follow-up suggestions plus lifestyle guidance.

Boundaries and Cautions

This skill is useful for health literacy, abnormal-value triage, and preparing information before a medical visit, but it cannot replace a doctor. If symptoms are mentioned alongside lab values, it does not infer causes; it advises taking the complete report to the relevant department for clinical judgment. Tumor markers, significant abnormalities, and emergency signs require separate emphasis on follow-up or prompt care. Reports contain personal health information, so users should be careful about uploading scope and storage.