Back to skills
extension
Category: Content & MediaNo API key required

section-detection

Detect clinical note sections (HPI, ROS, Assessment, Plan) using regex patterns and LLM topic segmentation. Generates ToC with accurate byte offsets.

personAuthor: jakexiaohubgithub

Section Detection Skill

Overview

Detects sections in clinical notes using hybrid approach: regex for explicit headers + LLM for inferred sections. Generates navigable Table of Contents with precise offsets.

When to Use

  • Generate Table of Contents for clinical notes
  • Detect section boundaries for navigation
  • Identify explicit and inferred sections

Installation

IMPORTANT: This skill has its own isolated virtual environment (.venv) managed by uv. Do NOT use system Python.

Initialize the skill's environment:

# From the skill directory
cd .agent/skills/section-detection
uv sync  # Creates .venv (no external dependencies, uses Python stdlib)

Usage

CRITICAL: Always use uv run to execute code with this skill's .venv, NOT system Python.

# From .agent/skills/section-detection/ directory
# Run with: uv run python -c "..."
from section_detection import SectionDetector

detector = SectionDetector(ollama_client)

sections = detector.detect_sections(clinical_note_text)

for section in sections:
    print(f"{section['title']}: {section['start_offset']}-{section['end_offset']}")
    print(f"  Explicit: {section['is_explicit']}, Confidence: {section['confidence']}")

Implementation

See section_detection.py.