gnn-supply-chain-risk

name	gnn-supply-chain-risk
description	Install or teardown the GNN Supply Chain Risk Intelligence solution. Usage: $sf-mleu-solutions:gnn-supply-chain-risk | $sf-mleu-solutions:gnn-supply-chain-risk teardown Triggers: gnn, graph neural network, supply chain risk, tier-2, concentration risk, bottleneck, pytorch geometric.
tools	["snowflake_sql_execute","snowflake_object_search","Bash","Read","Glob","Grep","WebFetch"]

GNN Supply Chain Risk Intelligence

Parse the action from $ARGUMENTS:

• If $ARGUMENTS is "install" or empty → run Install flow
• If $ARGUMENTS is "teardown" → run Teardown flow
• Otherwise → show usage help

Overview

• Industry: Manufacturing
• Database: SF_SOLUTIONS
• Schema: GNN_SUPPLY_CHAIN_RISK
• Features: Graph Neural Networks (PyTorch Geometric), SPCS GPU, Cortex Agent, Cortex Analyst, Semantic Model, Streamlit in Snowflake
• Role Required: ACCOUNTADMIN
• Optional: Snowpark Container Services (GPU compute pool), External Access Integration

Install

1. Locate the sf-mleu-solutions repository:

   • Check current working directory
   • If not found: git clone https://github.com/Snowflake-Labs/sf-mleu-solutions.git /tmp/sf-mleu-solutions

2. Read solutions/gnn-supply-chain-risk/manifest.json.

3. Query the current account info and present the installation plan:

   SELECT CURRENT_ORGANIZATION_NAME() AS ORG, CURRENT_ACCOUNT_NAME() AS ACCOUNT, CURRENT_REGION() AS REGION, CURRENT_ROLE() AS ROLE;
   

   Show to the user:

   Solution: GNN Supply Chain Risk Intelligence v1.0.0
   Industry: Manufacturing
   Database: SF_SOLUTIONS
   Schema:   GNN_SUPPLY_CHAIN_RISK
   Role:     ACCOUNTADMIN
   
   Target Account:
     Organization: <ORG>
     Account:      <ACCOUNT>
     Region:       <REGION>
     Current Role: <ROLE>
   
   What will be created:
     - 9 tables (VENDORS, MATERIALS, PURCHASE_ORDERS, BILL_OF_MATERIALS, TRADE_DATA, REGIONS, RISK_SCORES, PREDICTED_LINKS, BOTTLENECKS)
     - Demo data (~378 rows: 50 vendors, 26 materials, 120 POs, 150 trade records, 9 regions)
     - 4 analytics views
     - ANALYZE_RISK_SCENARIO SQL UDF
     - Semantic model (staged YAML for Cortex Analyst)
     - Cortex Agent (SUPPLY_CHAIN_RISK_AGENT)
     - RUN_RISK_SCORING stored procedure (NetworkX, no GPU required)
     - [Optional] GPU Compute Pool + External Access Integration (for PyTorch Geometric notebook)
   
   Proceed with installation?
   

4. Wait for user confirmation.

5. Read solutions/gnn-supply-chain-risk/scripts/setup.sql and execute in BATCHES:

   Batch 1 — Infrastructure (Section 1-2):

   USE ROLE ACCOUNTADMIN;
   CREATE DATABASE IF NOT EXISTS SF_SOLUTIONS;
   CREATE SCHEMA IF NOT EXISTS SF_SOLUTIONS.GNN_SUPPLY_CHAIN_RISK;
   USE DATABASE SF_SOLUTIONS;
   USE SCHEMA GNN_SUPPLY_CHAIN_RISK;
   CREATE WAREHOUSE IF NOT EXISTS SF_SOLUTIONS_WH WAREHOUSE_SIZE='LARGE' AUTO_SUSPEND=60 AUTO_RESUME=TRUE;
   USE WAREHOUSE SF_SOLUTIONS_WH;
   

   Then create stages (MODELS_STAGE, DATA_STAGE, SEMANTIC_MODELS).

   Batch 2 — Table DDL (Section 3): Execute all CREATE TABLE statements (VENDORS, MATERIALS, PURCHASE_ORDERS, BILL_OF_MATERIALS, TRADE_DATA, REGIONS + output tables RISK_SCORES, PREDICTED_LINKS, BOTTLENECKS).

   Batch 3 — Demo Data (Sections 4-9): Execute INSERT statements for VENDORS, MATERIALS, REGIONS, BILL_OF_MATERIALS, PURCHASE_ORDERS, TRADE_DATA. Use timeout_seconds: 300.

   Batch 4 — Analytics Views + UDF (Sections 10-11): Execute CREATE VIEW and CREATE FUNCTION statements.

   Batch 5 — Semantic Model (Section 12): Upload the semantic model YAML to stage. The YAML is inline in setup.sql — execute the PUT/copy statements from the script.

   Batch 6 — SPCS + EAI (Sections 13-14, OPTIONAL): Ask the user: "Do you want to create the GPU Compute Pool and External Access Integration? (Required for PyTorch Geometric notebook, NOT required for stored procedure risk scoring)"

   • If yes: execute Sections 13-14
   • If no: skip

   Batch 7 — Cortex Agent + Stored Procedure + Verification (Sections 15-17):

   CREATE OR REPLACE AGENT SF_SOLUTIONS.GNN_SUPPLY_CHAIN_RISK.SUPPLY_CHAIN_RISK_AGENT ...;
   CREATE OR REPLACE PROCEDURE SF_SOLUTIONS.GNN_SUPPLY_CHAIN_RISK.RUN_RISK_SCORING() ...;
   

6. Run the risk scoring stored procedure to populate output tables:

   CALL SF_SOLUTIONS.GNN_SUPPLY_CHAIN_RISK.RUN_RISK_SCORING();
   

   Use timeout_seconds: 600.

7. [MANDATORY — DO NOT SKIP] Retrieve and display the agent URL:

   SELECT
       'https://app.snowflake.com/' || LOWER(CURRENT_ORGANIZATION_NAME()) || '/' || LOWER(CURRENT_ACCOUNT_NAME())
           || '/#/agents/database/SF_SOLUTIONS/schema/GNN_SUPPLY_CHAIN_RISK/agent/SUPPLY_CHAIN_RISK_AGENT/details' AS AGENT_URL;
   

   Display to the user:

   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
   Supply Chain Risk Agent:
     <AGENT_URL>
   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
   

   This step is NON-OPTIONAL. The user must always see clickable URLs after install.

8. Show final summary:

   Installation complete: GNN Supply Chain Risk Intelligence v1.0.0
   
   Risk scoring completed — RISK_SCORES, PREDICTED_LINKS, and BOTTLENECKS tables populated.
   
   Next Actions:
   1. Open Snowsight > AI & ML > Agents > SUPPLY_CHAIN_RISK_AGENT
   2. Try: "What is our overall portfolio risk?"
   3. Try: "Which regions have the highest supply chain risk?"
   4. Try: "Simulate a vendor failure for V10006"
   
   Teardown: $sf-mleu-solutions:gnn-supply-chain-risk teardown
   

Teardown

If $ARGUMENTS is "teardown":

1. Confirm with user: "This will drop GNN_SUPPLY_CHAIN_RISK schema, GPU compute pool, and external access integration. Proceed?"
2. Read and execute solutions/gnn-supply-chain-risk/scripts/teardown.sql statement by statement.
3. Confirm: "GNN Supply Chain Risk Intelligence removed."

Next Actions

If the user asks "what next?", "what can I do?", or "how to customize":

Read and present the content from NEXT_ACTIONS.md (located in this skill's directory). Present the relevant section based on user intent:

• Just exploring → Quick Exploration section
• Wants GNN notebook → GNN Notebook section
• Wants to use own data → Customize with Your Data section

Usage Help

Usage:
  $sf-mleu-solutions:gnn-supply-chain-risk           — Install the solution
  $sf-mleu-solutions:gnn-supply-chain-risk teardown   — Remove the solution