Skip to content

Getting Started

Choose your path:

Quick Start (Local)

Run your first scan in 5 minutes using the CLI and Anthropic API. No cloud deployment needed.

Configuration

All environment variables, accounts.yaml reference, and AI provider options.

Your First Scan

What to expect from the output and how to interpret findings.

Prerequisites

  • Python 3.11+
  • Cloud credentials configured for your target cloud:
    • AWS: ~/.aws/credentials default profile, or AWS_PROFILE env var. AWS Resource Explorer must be enabled with an aggregator index in your primary region.
    • GCP: gcloud auth application-default login
    • Azure: az login
  • A Slack webhook URL — or set DRY_RUN=true to print to stdout instead
  • An AI provider (one of):
    • Anthropic API (easiest for local dev): set ANTHROPIC_API_KEY
    • AWS Bedrock: uses your IAM role automatically (AWS only)
    • Vertex AI: uses ADC automatically (GCP only)
    • Azure OpenAI: uses managed identity automatically (Azure only)

GCP and Azure

argus scan --cloud gcp and argus scan --cloud azure work locally with credentials configured. For automated weekly scans, deploy the Cloud Run Job or Azure Function (see the Deployment section).