hacs-tools 0.3.1

Core tools and utilities for HACS (Healthcare Agent Communication Standard)

HACS Tools

25+ Healthcare tools for AI agents via Model Context Protocol

Production-ready tools for clinical workflows, resource management, and healthcare AI operations.

🛠️ Tool Categories

🔍 Resource Discovery & Development (5 tools)

• discover_hacs_resources - Explore healthcare resource schemas with metadata
• analyze_resource_fields - Field analysis with validation rules
• compare_resource_schemas - Schema comparison and integration
• create_clinical_template - Generate clinical workflow templates
• create_model_stack - Compose complex data structures

📋 Record Management (8 tools)

• create_hacs_record / get_hacs_record_by_id / update_hacs_record / delete_hacs_record - Full CRUD
• validate_hacs_record_data - Comprehensive validation
• list_available_hacs_resources - Resource schema catalog
• find_hacs_records - Advanced semantic search
• search_hacs_records - Filtered record search

🧠 Memory Management (5 tools)

• create_memory - Store episodic/procedural/executive memories
• search_memories - Semantic memory retrieval
• consolidate_memories - Merge related memories
• retrieve_context - Context-aware memory access
• analyze_memory_patterns - Usage pattern analysis

✅ Validation & Schema (3 tools)

• get_hacs_resource_schema - JSON schema exploration
• create_view_resource_schema - Custom view creation
• suggest_view_fields - Intelligent field suggestions

🎨 Advanced Tools (3 tools)

• optimize_resource_for_llm - LLM-specific optimizations
• version_hacs_resource - Resource versioning and tracking

📚 Knowledge Management (1 tool)

• create_knowledge_item - Clinical guidelines and protocols

📦 Installation

pip install hacs-tools

🚀 Quick Start

import requests

def call_hacs_tool(tool_name, arguments):
    """Call HACS MCP tools"""
    response = requests.post('http://localhost:8000/', json={
        "jsonrpc": "2.0",
        "method": "tools/call",
        "params": {
            "name": tool_name,
            "arguments": arguments
        },
        "id": 1
    })
    return response.json()

# Create patient record
patient_result = call_hacs_tool("create_hacs_record", {
    "resource_type": "Patient",
    "resource_data": {
        "full_name": "John Smith",
        "birth_date": "1980-05-15",
        "gender": "male"
    }
})

# Store clinical memory
memory_result = call_hacs_tool("create_memory", {
    "content": "Patient reports improved symptoms after treatment",
    "memory_type": "episodic",
    "importance_score": 0.8
})

# Search for related memories
search_result = call_hacs_tool("search_memories", {
    "query": "treatment response",
    "limit": 5
})

🏥 Healthcare Workflows

Clinical Assessment

# Generate assessment template
template = call_hacs_tool("create_clinical_template", {
    "template_type": "assessment",
    "focus_area": "cardiology",
    "complexity_level": "standard"
})

# Create knowledge item
knowledge = call_hacs_tool("create_knowledge_item", {
    "title": "AHA Guidelines 2024",
    "content": "New recommendations for hypertension management",
    "knowledge_type": "guideline"
})

Resource Discovery

# Discover available models
models = call_hacs_tool("discover_hacs_resources", {
    "category_filter": "clinical",
    "include_examples": True
})

# Get schema for specific model
schema = call_hacs_tool("get_hacs_resource_schema", {
    "resource_type": "Patient",
    "include_validation_rules": True
})

🔗 Integration

HACS Tools integrate with:

• MCP Protocol - Standard tool calling interface
• LangGraph - AI agent workflows
• PostgreSQL - Persistent healthcare data storage
• Healthcare Systems - FHIR-compliant data exchange

📊 Performance

• Tool Execution: <200ms average response time
• Memory Search: <100ms for semantic queries
• Resource Creation: <50ms for standard resources
• Validation: <10ms for schema validation

📄 License

Apache-2.0 License - see LICENSE for details.