Discovery Tools and Guidance

Advanced tools for finding the right commands and understanding how to use them.

Command Recommendations (hana_recommend)

Get command suggestions based on what you want to do in plain English.

How It Works

The recommendation system matches your intent to commands using:

1. Pattern Matching - Keywords you use
2. Confidence Scoring - How well each command matches
3. Example Parameters - Ready-to-use parameter templates

Basic Usage

Input: Natural language description of what you want to do

{
  "intent": "find duplicate rows"
}

Output: Top 5 matching commands with confidence and reasoning

{
  "recommendations": [
    {
      "rank": 1,
      "command": "duplicateDetection",
      "confidence": "high",
      "reasoning": "Designed specifically for finding duplicate records",
      "exampleParameters": {
        "table": "MY_TABLE",
        "schema": "SALES",
        "checkColumns": "EMAIL,PHONE"
      }
    },
    {
      "rank": 2,
      "command": "dataProfile",
      "confidence": "medium",
      "reasoning": "Can identify duplicate patterns and distribution",
      "exampleParameters": {
        "table": "MY_TABLE",
        "schema": "SALES"
      }
    }
  ]
}

Common Intent Patterns

Intent	Recommended Command
"List tables"	hana_tables, hana_schemas
"Find duplicates"	hana_duplicateDetection, hana_dataProfile
"Import data"	hana_import, hana_tableCopy, hana_dataSync
"Check data quality"	hana_dataValidator, hana_dataProfile
"Compare tables"	hana_compareData, hana_dataDiff
"Export data"	hana_export, hana_tableCopy
"Analyze performance"	hana_memoryAnalysis, hana_expensiveStatements
"Verify connection"	hana_status, hana_healthCheck
"Find missing columns"	hana_compareSchema, hana_inspectTable

Consolidated Search (hana_search)

Comprehensive search across all resources in the MCP Server.

Search Scope

The smart search looks across:

1. Commands (150+)

   • Command names
   • Descriptions
   • Tags and categories
   • Use cases

2. Workflows (20+)

   • Multi-step task sequences
   • Workflow names and descriptions
   • Task phases

3. Examples (40+)

   • Real-world scenarios
   • Parameter examples
   • Usage descriptions

4. Presets (30+)

   • Parameter templates
   • Preset names
   • Use case descriptions

Usage Examples

Example 1: Find Import Commands

{
  "query": "import CSV",
  "scope": ["commands"],
  "limit": 5
}

Result:

{
  "results": [
    {
      "type": "command",
      "name": "import",
      "relevance": 99,
      "description": "Import data from CSV file",
      "category": "data-operations"
    },
    {
      "type": "command", 
      "name": "tableCopy",
      "relevance": 78,
      "description": "Copy table data, supports CSV format",
      "category": "data-operations"
    }
  ]
}

Example 2: Find Data Quality Workflows

{
  "query": "data quality validation",
  "scope": ["workflows", "examples"],
  "limit": 10
}

Result:

{
  "results": [
    {
      "type": "workflow",
      "name": "data-quality-check",
      "relevance": 95,
      "phase": 1,
      "steps": 5
    },
    {
      "type": "example",
      "command": "dataValidator",
      "scenario": "Comprehensive Data Quality Check",
      "relevance": 88
    },
    {
      "type": "preset",
      "command": "dataProfile",
      "name": "quality-analysis",
      "relevance": 82
    }
  ]
}

Ranking Algorithm

Results are scored based on:

• Exact match - 100 points
• Contains phrase - 50 points
• Contains all words - 30 points
• Per word match - 10 points
• Start of text bonus - 20 points

Results are sorted by total relevance score.

Quick Start Guide (hana_quickstart)

Perfect for users new to HANA CLI and the MCP Server.

The 6 Essential Commands

The quick start teaches you these commands in order:

1. hana_status - Verify connection and current user

   # Shows:
   # - Current user
   # - Connected database
   # - Current schema
   # - Version information

   Why: Confirms everything is set up correctly

2. hana_version - Check database version

   # Shows HANA version and build number

   Why: Understand database capabilities

3. hana_schemas - List available schemas

   # Shows all schemas you can access

   Why: See available data sources

4. hana_tables - List tables in a schema

   hana_tables --schema SALES
   # Shows all tables in SALES schema

   Why: Find the data you need to work with

5. hana_inspectTable - View table structure

   hana_inspectTable --table CUSTOMERS --schema SALES
   # Shows columns, types, constraints

   Why: Understand what data is available

6. hana_healthCheck - Check system health

   # Shows warnings and critical issues

   Why: Ensure database is running properly

Next Commands to Learn

After the quick start, explore:

• Data Analysis: dataProfile, dataValidator, duplicateDetection
• Data Operations: import, export, dataSync
• Schema Tools: compareSchema, schemaClone, inspectTable
• System Tools: memoryAnalysis, expensiveStatements, recommendations

Conversation Templates (hana_conversation_templates)

Pre-built conversation flows for common tasks.

Available Templates

1. Data Exploration (15-30 minutes)

Goal: Understand database structure and data

Phases:

1. Verify connection
2. Check version and system info
3. List available schemas
4. Explore tables in key schemas
5. Profile data quality

Commands:

• hana_status
• hana_version
• hana_systemInfo
• hana_schemas
• hana_tables
• hana_inspectTable
• hana_dataProfile

Tips:

• Start with SYSTEM schema for system tables
• Look for SALES, CUSTOMER, PRODUCT schemas
• Profile small tables first to understand data

2. Troubleshooting (20-40 minutes)

Goal: Diagnose and fix issues

Phases:

1. Check system health
2. Verify connectivity
3. Test permissions
4. Analyze resource usage
5. Identify bottlenecks

Commands:

• hana_healthCheck
• hana_status
• hana_inspectUser
• hana_memoryAnalysis
• hana_expensiveStatements
• hana_recommendations

Tips:

• Run healthCheck first
• Check user roles and privileges
• Top queries often cause performance issues
• Review system alerts

3. Data Migration (30-60 minutes)

Goal: Move data between sources

Phases:

1. Validate source schema
2. Prepare target schema
3. Export source data
4. Import to target
5. Verify migration

Commands:

• hana_inspectTable
• hana_compareSchema
• hana_export
• hana_import
• hana_compareData
• hana_dataValidator

Tips:

• Always do dry-run first
• Start with small tables
• Validate after import
• Check data quality

4. Performance Tuning (30-60 minutes)

Goal: Optimize database performance

Phases:

1. Establish baseline
2. Identify hotspots
3. Analyze indexes
4. Review recommendations
5. Implement improvements

Commands:

• hana_memoryAnalysis
• hana_tableHotspots
• hana_indexTest
• hana_recommendations
• hana_expensiveStatements

Tips:

• Large tables cause most issues
• Monitor memory usage trends
• Test index effectiveness
• Compare before/after metrics

5. Security Audit (20-40 minutes)

Goal: Review and secure database access

Phases:

1. Inventory users
2. Review roles and privileges
3. Check inactive accounts
4. Audit recent access
5. Identify issues

Commands:

• hana_users
• hana_inspectUser
• hana_roles
• hana_auditLog
• hana_replicationStatus

Tips:

• Document all user accounts
• Review DBA privileges
• Look for inactive accounts
• Check sensitive table access

Using a Template

Step 1: Get Template

{
  "templateId": "data-exploration"
}

Step 2: Review Steps

Each step includes:

• Purpose and goal
• Commands to run
• Expected outcomes
• Tips for success

Step 3: Follow Guided Workflow

• Run each command in sequence
• Review results
• Proceed to next step
• Skip optional steps if needed

Context-Aware Suggestions

After Command Execution

The system automatically suggests useful next steps:

After hana_status:

✅ Connected successfully

🔄 Suggested Next Steps:
1. Explore available schemas
   → Use: hana_schemas
   → Helps: Understand available data

2. Check system health
   → Use: hana_healthCheck
   → Helps: Verify system is stable

3. View version information  
   → Use: hana_version
   → Helps: Understand capabilities

After hana_tables:

✅ Listed 47 tables

🔄 Suggested Next Steps:
1. Inspect interesting table structures
   → Use: hana_inspectTable
   → Parameter: table="CUSTOMERS"

2. Profile table data quality
   → Use: hana_dataProfile
   → Parameter: table="CUSTOMERS"

3. Find duplicate rows
   → Use: hana_duplicateDetection
   → Parameter: table="CUSTOMERS"

Output-Based Tips

Tips appear based on command results:

When import has errors:

⚠️ Import completed with 5 errors

📌 Suggested Actions:
• Try dryRun:true to preview before actual import
• Use skipWithErrors:true to continue on error
• Check error log for details
• Review data validation with dataValidator

When memory usage is high:

⚠️ Memory usage at 85% capacity

📌 Suggested Actions:
• Review expensiveStatements to find heavy queries
• Check tableHotspots for concentrated usage
• Implement recommendations
• Monitor memoryAnalysis trends

Best Practices for Discovery

1. Start with Recommendations

For any task, start with hana_recommend:

{
  "intent": "what you want to do"
}

This ensures you're using the right command.

2. Review Examples

Before running a command, check examples:

{
  "command": "import"
}

Learn from real-world usage patterns.

3. Use Parameter Presets

Let presets guide your parameters:

{
  "command": "import"
}

Get templates for different scenarios.

4. Follow Templates

For complex tasks, use conversation templates:

• Data exploration
• Troubleshooting
• Data migration
• Performance tuning
• Security audits

5. Interpret Results

Use result interpretation for insights:

{
  "command": "dataProfile",
  "result": "command output"
}

Get AI-friendly analysis and recommendations.

Workflow Discovery

List Available Workflows

{
  "action": "list"
}

Returns all 20+ pre-built workflows.

Get Workflow Details

{
  "workflowId": "data-quality-check"
}

Returns complete steps, parameters, and examples.

Search Workflows

{
  "query": "data validation",
  "scope": "workflows"
}

Find workflows matching your needs.

Execute Workflow

{
  "action": "execute",
  "workflowId": "data-quality-check",
  "parameters": {
    "schema": "SALES",
    "table": "CUSTOMERS"
  }
}

Runs complete multi-step workflow.

Examples for Commands

Get Examples

{
  "command": "import"
}

Returns 5+ real-world scenarios:

• Quick CSV import
• Large file with error handling
• Streaming mode
• Bulk insert
• Error validation

What Examples Include

• Complete parameter set
• Scenario description
• Expected output
• Tips and best practices
• Common issues and solutions

Using Examples in Your Work

1. Find relevant scenario
2. Copy parameter template
3. Customize for your data
4. Run with dryRun first
5. Execute when confident

Next Steps

• Advanced Features - Workflows and interpretation
• Features Overview - All available capabilities
• Prompts and Resources - MCP resources and prompts