Skip to main content

Usage Examples

Real-world examples of using SolarSens MCP with Claude.

Daily Operations​

Morning Check​

You ask:

"Good morning! How did my solar plants perform yesterday?"

Claude will:

  1. Call fetch_operation_overview for yesterday's date
  2. Summarize the performance across all plants
  3. Highlight any issues or anomalies

Example response:

Yesterday's performance summary for your 5 plants:

✅ Top Performers:

  • Ho Chi Minh City Plant 1: 12,500 kWh (PR: 85%)
  • Bangkok Solar 2: 20,100 kWh (PR: 87%)

⚠️ Attention Needed:

  • Singapore Plant 3: 5,200 kWh (PR: 68% - below expected)

Overall portfolio: 62,800 kWh generated

Troubleshooting​

Investigate Underperformance​

You ask:

"The Singapore Plant 3 seems low. Can you analyze the last 7 days?"

Claude will:

  1. Call analyze_plant_history with plant name and date range
  2. Identify trends and patterns
  3. Provide insights

Example response:

Analysis for Singapore_Plant_3 (Last 7 days):

Trend: Performance declining since Oct 25

  • Oct 25: PR 82% (normal)
  • Oct 26-28: PR 75-78% (slight decline)
  • Oct 29-31: PR 68-70% (significant drop)

Possible causes:

  • Availability dropped from 98% to 92%
  • Suggests equipment issues or maintenance needed

Portfolio Management​

Compare Plants​

You ask:

"Which of my plants had the best performance ratio last month?"

Claude will:

  1. Call list_plants to get all plants
  2. Call analyze_plant_history for each plant for the date range
  3. Compare and rank

Example response:

September 2024 Performance Rankings:

🥇 Bangkok Solar 2: 86.5% avg PR 🥈 Ho Chi Minh City Plant 1: 84.2% avg PR 🥉 Thailand Plant 5: 83.8% avg PR

Recommendations:

  • Bangkok Solar 2 is your best performer - study its maintenance schedule
  • Consider applying similar practices to lower-performing plants

Reporting​

Weekly Summary​

You ask:

"Generate a weekly summary report for all plants"

Claude will:

  1. Fetch data for the past 7 days
  2. Calculate key metrics
  3. Format a comprehensive report

Example response:

Weekly Report: Oct 25-31, 2024

Portfolio Overview:

  • Total Energy: 438,500 kWh
  • Average PR: 81.4%
  • Fleet Availability: 96.2%

Plant Breakdown:

PlantEnergy (kWh)Avg PRAvailability
Bangkok Solar 2140,70087%99%
Ho Chi Minh City 187,50085%98%
Singapore Plant 336,40072%92%
...

Action Items:

  • Investigate Singapore Plant 3 availability issues
  • Schedule maintenance for plants with PR < 75%

Advanced Queries​

Date Range Analysis​

You ask:

"Compare this month vs last month for Bangkok Solar 2"

Claude will:

  1. Analyze current month data
  2. Analyze previous month data
  3. Calculate differences and trends

Specific Metrics​

You ask:

"What was the availability trend for all plants in Q3?"

Claude will:

  1. Fetch Q3 data for all plants
  2. Calculate availability metrics
  3. Show trends and insights

Predictive Insights​

You ask:

"Based on the last 30 days, which plants might need attention soon?"

Claude will:

  1. Analyze 30-day performance trends
  2. Identify declining patterns
  3. Provide proactive recommendations

Tips for Best Results​

  1. Be Specific with Dates

    • ✅ "Show me data for October 15-20"
    • ❌ "Show me recent data"

  2. Use Plant Names

    • ✅ "Analyze Ho Chi Minh City plant"
    • ✅ "Analyze plant with ID plant_123"
    • Fuzzy matching works: "ho chi" finds "Ho_Chi_Minh_City_1"

  3. Ask Follow-up Questions

    • Claude remembers context in the conversation
    • "Now show me the same for Bangkok plant" works after analyzing one plant

  4. Request Comparisons

    • "Compare X vs Y"
    • "Which plant performed best?"
    • "Rank plants by availability"

Next Steps​