RepoSnowflake (Arctic)Snowflake (Arctic)published Dec 9, 2025seen 5d

Snowflake-Labs/si-cybersecurity-challenge

PLpgSQL

Open original ↗

Captured source

source ↗

Snowflake-Labs/si-cybersecurity-challenge

Language: PLpgSQL

License: Apache-2.0

Stars: 10

Forks: 1

Open issues: 0

Created: 2025-12-09T06:00:07Z

Pushed: 2025-12-09T13:41:38Z

Default branch: main

Fork: no

Archived: no

README:

Snowflake Intelligence CTF - Cybersecurity Challenge

A Capture The Flag (CTF) challenge that demonstrates Snowflake's Intelligence capabilities using cybersecurity log analysis with Cortex Agents and Semantic Models.

Challenge 1

Someone is trying to exfiltate data from a Snowflake account. However they are trying to avoid detection. Use Snowflake inteligence to uncover evidence. The flag will be obvious and have a format like ctf{FLAG}

Deployment Steps

Prerequisites

Before deploying this challenge, ensure you have:

1. Snowflake Account with access to:

  • ACCOUNTADMIN role (or equivalent privileges)
  • Snowflake Cortex features enabled
  • Snowflake Intelligence enabled

2. Enable Snowflake Intelligence:

Step 1: Run the Setup Script

1. Open Snowsight (https://app.snowflake.com) 2. Navigate to Projects > Worksheets 3. Create a new SQL worksheet 4. Copy the entire contents of setup_snowflake.sql from this repository 5. Paste it into the worksheet 6. Click Run All to execute the script

[Screenshot: Snowsight worksheet with setup_snowflake.sql]

The script will automatically:

  • Create the SI_CYBERSECURITY_CHALLENGE_DB database
  • Create the SI_CYBERSECURITY_CHALLENGE_WH warehouse
  • Create NETWORK_LOGS and QUERY_LOGS tables
  • Load sample data from GitHub
  • Create a semantic model named SI_CYBERSECURITY_CHALLENGE

Expected Output: The script should complete successfully with messages confirming data load and semantic model creation.

Create the Cortex Agent

Now that your data and semantic model are set up, create a Cortex Agent to enable natural language queries.

Step 1: Navigate to Agents

1. In Snowsight, go to AI & ML > Agents 2. Click + Create agent

Step 2: Configure Basic Settings

1. Select the following options:

  • Platform integration: Leave the box for "Create this agent for Snowflake Intelligence" checked
  • Agent object name: SI_CYBERSECURITY_ANALYST
  • Display name: Snowflake Intelligence Challenge Agent

2. Click Create agent

![Create Agent](assets/create_agent.png)

Step 3: Edit Agent Configuration

After creation, open the agent and press the "edit" button

About -> Description

An AI agent for analyzing cybersecurity logs including network traffic and query patterns.
Use this agent to investigate security incidents, identify anomalies, and analyze user behavior.

Orchestration -> Orchestration instructions

You are a cybersecurity analyst assistant specializing in log analysis.

Your role is to:
- Analyze network logs and query logs to identify patterns and anomalies
- Help users understand security-related metrics and trends
- Provide clear, actionable insights about potential security issues
- Explain findings in a professional yet accessible manner

#### Orchestration -> Response instructions

When answering questions:
- Be specific and cite the data sources
- Highlight unusual patterns or outliers
- Suggest follow-up questions when relevant
- Prioritize security-critical information

Step 4: Add Cortex Analyst Tool

1. Scroll to the Tools section 2. Click + Add next to Cortex Analyst

3. Configure the Cortex Analyst tool:

  • Schema: SI_CYBERSECURITY_CHALLENGE_DB.PUBLIC
  • Semantic view: Select SI_CYBERSECURITY_CHALLENGE from the dropdown
  • Name: SI_Cybersecurity_Semantic_Model
  • Leave the rest of the options as default

4. Generate the description with cortex 5. Click Add to save the tool

![Cortex Analyst Configuration](assets/analyst_config.png)

Step 5: Save and Test

1. Click Save in the top right corner to save all agent configurations 2. Test the agent by asking one of the sample questions in the chat interface 3. Verify that the agent responds with data from your semantic model

Step 6: Access the Agent

1. Go to ai.snowflake.com 2. Click on "Agents" in the left navigation menu 3. Find and select your newly created cybersecurity agent from the list 4. The agent is now ready to analyze your security logs and answer questions

License

Copyright (c) Snowflake Inc. All rights reserved. Licensed under the Apache 2.0 license.

Notability

notability 2.0/10

Low stars, routine repo