Morning Market Scan

2. Pull USDA Grain Data for ALL Crops

Run each crop check individually:

Corn (ND):

curl -s --max-time 20 "https://crop-intel-api-production.up.railway.app/api/usda/grain-report?commodity=Corn&state=ND" | python3 -c "
import json,sys
d=json.load(sys.stdin)
s=d.get('data',{}).get('summary',{})
print(f'CORN | Source: {d.get(\"source\")} | Avg: \${s.get(\"avgPrice\",0):.2f}/bu | Basis: {s.get(\"avgBasis\",0):.0f}¢ | Bids: {s.get(\"bidCount\",0)} | Date: {s.get(\"reportDate\",\"?\")}')
"

Soybeans (ND):

curl -s --max-time 20 "https://crop-intel-api-production.up.railway.app/api/usda/grain-report?commodity=Soybeans&state=ND" | python3 -c "
import json,sys
d=json.load(sys.stdin)
s=d.get('data',{}).get('summary',{})
print(f'SOYBEANS | Source: {d.get(\"source\")} | Avg: \${s.get(\"avgPrice\",0):.2f}/bu | Basis: {s.get(\"avgBasis\",0):.0f}¢ | Bids: {s.get(\"bidCount\",0)}')
"

Wheat (ND):

curl -s --max-time 20 "https://crop-intel-api-production.up.railway.app/api/usda/grain-report?commodity=Wheat&state=ND" | python3 -c "
import json,sys
d=json.load(sys.stdin)
s=d.get('data',{}).get('summary',{})
print(f'WHEAT | Source: {d.get(\"source\")} | Avg: \${s.get(\"avgPrice\",0):.2f}/bu | Basis: {s.get(\"avgBasis\",0):.0f}¢ | Bids: {s.get(\"bidCount\",0)}')
"

Sunflowers:

curl -s --max-time 20 "https://crop-intel-api-production.up.railway.app/api/usda/sunflower-report" | python3 -c "
import json,sys
d=json.load(sys.stdin)
print(f'SUNFLOWERS | Source: {d.get(\"source\")} | Degraded: {d.get(\"degraded\")}')
"

3. Check Best Opportunities Per Crop

For each crop, pull the buyer network and sort by net price:

curl -s --max-time 15 "https://crop-intel-api-production.up.railway.app/api/buyers?scope=all&crop=Yellow%20Corn" | python3 -c "
import json,sys
d=json.load(sys.stdin)
buyers=d.get('data',[])
# Sort by cash price descending
with_price = [b for b in buyers if b.get('cashBid') or b.get('postedBasis')]
print(f'=== YELLOW CORN: {len(buyers)} buyers ===')
for b in buyers[:5]:
    phone = b.get('facilityPhone','No Phone')
    rc = b.get('railConfidence',0) or 0
    rail = 'BNSF' if rc >= 70 else 'Likely' if rc >= 40 else 'Truck'
    print(f'  {b[\"name\"]} ({b[\"state\"]}) | Phone: {phone} | Rail: {rail}')
"

Repeat for Soybeans, Wheat, Sunflowers.

4. Generate AI Insights for Top Buyers

For each promising buyer (score > 60), get Gemini explanation:

curl -s --max-time 30 -X POST "https://crop-intel-api-production.up.railway.app/api/ai/buyer-intel" \
  -H "Content-Type: application/json" \
  -d '{
    "crop":"Yellow Corn",
    "buyerData":{
      "name":"BUYER_NAME",
      "type":"BUYER_TYPE",
      "city":"BUYER_CITY",
      "state":"BUYER_STATE",
      "netPrice":NET_PRICE,
      "benchmarkPrice":BENCHMARK,
      "railConfidence":RAIL_CONFIDENCE,
      "hasPhone":true,
      "freightCost":FREIGHT
    },
    "withExplanation":true
  }'

5. Generate Morning Summary

Compile findings into a clear morning briefing:

• Which crop has the best net price vs local benchmark today?
• Top 3 buyers to call for each active crop
• Any unusual market movements (basis changes > 5¢)
• Rail freight updates

6. Live Verification

Open the app at https://crop-intel-usa.vercel.app/ and verify:

• Benchmark price matches USDA data
• Live Market Quotes show correct basis/cash
• Buyer network count matches API response
• Crop switching works (Corn → Soybeans → Wheat → Sunflowers)
• AI intel scores are populating

Output Format

Present findings as a table:

Farm Context

• Origin: Campbell, MN (Zip 56522)
• Rail Access: BNSF mainline
• Local Benchmark: Hankinson Renewable Energy (corn), ADM Enderlin (sunflowers)
• Freight: $0.30/bu truck to Hankinson, BNSF Tariff 4022 for rail