Glimpse

Execute code snippets in secure, isolated environments via API

Glimpse is a lightweight API service that securely runs user-submitted code in sandboxed environments. Demonstrates secure code execution patterns with safety constraints and multiple isolation strategies.

Features

• Language support: Python, JS, C, Go, Kotlin (more to come)
• Isolation: Ephemeral containers or serverless functions per request
• Security controls:
  • Network access restrictions
  • Filesystem operation blocking
  • 30-second timeout enforcement
• Architecture options:
  • AWS Lambda deployment for serverless operation
  • Docker-based container pool for local development

API Usage

Endpoint: https://glimpse-7eir.onrender.com/run-code-lambda

Example request:

fetch("https://glimpse-7eir.onrender.com/run-code-lambda", {
  method: "POST",
  headers: { "Content-Type": "application/json" },
  body: JSON.stringify({
    language: "py",
    code: 'print("Hello from Glimpse!")',
    input: ""
  }),
})
.then(response => response.json())
.then(data => console.log(data));

Example response:

{
  "statusCode": 200,
  "body": {
    "output": "Hello from Glimpse!\n",
    "error": null,
    "executionTime": null
  }
}

Deployment

AWS Lambda Configuration

Set environment variables:

AWS_ACCESS_KEY_ID=your_key
AWS_SECRET_ACCESS_KEY=your_secret
AWS_DEFAULT_REGION=your_region
LAMBDA_FUNCTION_NAME=your_lambda_function

Follow infrastructure deployment guide in /docs/infra.markdown

Start API server:

uvicorn api-lambda:app --host 0.0.0.0 --port 8000

Docker-based Execution

Build container environment:

make build  # or ./setup.sh

Launch service:

uvicorn api-docker:app --host 0.0.0.0 --port 8000

Security Constraints

• Maximum execution duration: 30 seconds
• Rate limiting: 1000 requests/hour per IP
• Restricted system access:
  • No persistent storage
  • No external network connectivity
  • Input limited to API parameters

Architecture

Implements two isolation strategies:

Serverless Execution

• Utilizes AWS Lambda for automatic scaling and isolation through AWS Firecracker microVMs

Container Pooling

• Maintains pre-warmed Docker containers for rapid execution in development environments

Core technologies:

• FastAPI (Python web server)
• Docker (Container runtime)
• AWS Lambda (Serverless compute)

License

MIT License