.env.example to .env and configure the variables below.
Quick Start
For local development:Core Configuration
Environment mode:
dev, staging, or prodDatabase
PostgreSQL database configuration (required).PostgreSQL username
PostgreSQL password. Auto-generated by
make init.PostgreSQL database name
PostgreSQL host (Docker service name or hostname)
PostgreSQL port
Redis
Redis is used for Celery task queue and caching.Redis connection URL
Object Storage
S3-compatible object storage configuration. See Storage Configuration for details.S3 bucket name
S3 endpoint URL. Default uses SeaweedFS (included in docker-compose).For AWS S3, omit this variable. For other providers (Cloudflare R2, MinIO, etc.), set the endpoint URL.
S3 access key ID
S3 secret access key
S3 region
Enable SSL/TLS for storage connections
Verify SSL certificates. Set to
true in production for security.Enable Redis caching for storage file listings
Cache TTL in seconds
URLs & Networking
Frontend URL (for internal use)
Backend URL (for internal container-to-container communication)
Backend URL accessible from the browser. Injected at container startup.
WebSocket URL for chatbot. Injected at container startup.
Internal chatbot service URL
NEXT_PUBLIC_* variables are injected at container startup, not baked at build time. To change them:- Update
.env - Recreate the frontend container:
docker compose -f docker-compose.prod-local.yml up -d frontend
Authentication & Security
Flask session secret key. Auto-generated by
make init.Flask API port
Auth.js secret key. Auto-generated by
make init.HashiCorp Vault root token. Auto-generated by
make init.See Vault Secrets for details.Vault server address
Vault KV v2 engine mount point
Base path for user secrets in Vault
LLM & AI Services
At least one LLM provider API key is required. See LLM Providers for details.OpenRouter API key (recommended). Get from https://openrouter.ai/keysSupports multiple models from various providers through a single API.
OpenAI API key. Get from https://platform.openai.com/api-keys
Anthropic API key. Get from https://console.anthropic.com/
Google AI (Gemini) API key. Get from https://aistudio.google.com/app/apikey
Default LLM provider:
openrouter, openai, anthropic, or googleMaximum recursion depth for agent workflows
Web Search
SearxNG internal URL
SearxNG public URL
SearxNG secret key. Auto-generated by
make init.AI Features
Enable cost optimization for root cause analysis
Disable “thinking” mode for Gemini models
Cloud Provider Integrations
Optional cloud provider credentials for resource discovery and management.Google Cloud Platform
GCP OAuth client ID
GCP OAuth client secret
Amazon Web Services
AWS access key ID
AWS secret access key
AWS default region
Cloud Provider Cache
Applies to all cloud providers (GCP, AWS, Azure).Enable caching for cloud provider API responses
Cache TTL in seconds (1 hour)
Verify CLI identity for cloud provider connections
Cache cloud provider tokens in Redis
Third-Party Integrations
Optional integrations with external services.GitHub OAuth
GitHub OAuth client ID
GitHub OAuth client secret
Slack
Enable Slack integration
Slack OAuth client ID
Slack OAuth client secret
Slack signing secret
PagerDuty
Enable PagerDuty OAuth integration
PagerDuty OAuth client ID
PagerDuty OAuth client secret
Bitbucket
Enable Bitbucket integration
Bitbucket OAuth client ID
Bitbucket OAuth client secret
OVH (Multi-Region)
Enable OVH integration
OVH Europe client ID
OVH Europe client secret
OVH Canada client ID
OVH Canada client secret
OVH US client ID
OVH US client secret
Scaleway
Enable Scaleway integration
Confluence
Enable Confluence integration
Confluence OAuth client ID
Confluence OAuth client secret
Dynatrace
Enable Dynatrace integration
BigPanda
Enable BigPanda integration
ThousandEyes
Enable ThousandEyes integration
Email (SMTP)
SMTP server hostname
SMTP server port
SMTP username
SMTP password
Email sender address
Email sender name
Kubernetes & Pod Isolation
Configuration for running Aurora agents in isolated Kubernetes pods.Enable isolated pod execution for terminal and chatbot agents
Kubernetes namespace for terminal pods
Container image for terminal pods
Time-to-live for terminal pods (seconds)
Runtime class for terminal pods (e.g.,
gvisor, kata-containers)Time-to-live for chatbot pods (seconds)
Schedule pods on untrusted nodes (with appropriate taints/tolerations)
Helm chart URL for kubectl agent deployment
Weaviate (Vector Database)
Weaviate host (Docker service name or hostname)
Weaviate HTTP port
Weaviate gRPC port
Memgraph (Graph Database)
Memgraph host (Docker service name or hostname)
Memgraph Bolt protocol port
Memgraph username
Memgraph password
Resource discovery interval (hours)
Rate Limiting
Enable API rate limiting
Token to bypass rate limits (for testing/admin)
Include rate limit headers in API responses
Development
ngrok tunnel URL for local development (webhooks, OAuth callbacks)