.example.env

# copy this file to .env and fill in the values

# path to google cloud credentials file
ANOMSTACK_GOOGLE_APPLICATION_CREDENTIALS=
# can also be set as json string
ANOMSTACK_GOOGLE_APPLICATION_CREDENTIALS_JSON=

# gcp related env vars
ANOMSTACK_GCP_PROJECT_ID=

# snowflake related env vars
ANOMSTACK_SNOWFLAKE_ACCOUNT=
ANOMSTACK_SNOWFLAKE_USER=
ANOMSTACK_SNOWFLAKE_PASSWORD=
ANOMSTACK_SNOWFLAKE_WAREHOUSE=

# aws related env vars
ANOMSTACK_AWS_ACCESS_KEY_ID=
ANOMSTACK_AWS_SECRET_ACCESS_KEY=

# admin authentication (for Fly.io deployment and admin access)
ANOMSTACK_ADMIN_USERNAME=admin
ANOMSTACK_ADMIN_PASSWORD=anomstack2024

# local duckdb path for testing/dev quickstart
# For local development (outside containers)
# ANOMSTACK_DUCKDB_PATH=tmpdata/anomstack-duckdb.db
# For container deployments (Docker Compose, Fly.io, etc.)
ANOMSTACK_DUCKDB_PATH=/data/anomstack.db
# example using motherduck
# https://motherduck.com/docs/getting-started/connect-query-from-python/choose-database/
# ANOMSTACK_DUCKDB_PATH=md:anomstack

# local sqlite path for testing/dev quickstart
ANOMSTACK_SQLITE_PATH=tmpdata/anomstack-sqlite.db
# example using turso
# https://docs.turso.tech/sdk/python/quickstart
# ANOMSTACK_SQLITE_PATH=libsql://<your-database-url>.turso.io

# table id to store metrics in
ANOMSTACK_TABLE_KEY=tmp.metrics

# path for where to store models
# local path
ANOMSTACK_MODEL_PATH=local://./tmp/models
# for gcs bucket
# ANOMSTACK_MODEL_PATH=gs://your-bucket/models
# for s3 bucket
# ANOMSTACK_MODEL_PATH=s3://your-bucket/models

# home directory for anomstack (defaults to current working directory)
# used for docker volume mounts in dagster_docker.yaml
ANOMSTACK_HOME=.

# slack bot token for alerting
# ANOMSTACK_SLACK_BOT_TOKEN=
# slack channel for alerting
# ANOMSTACK_SLACK_CHANNEL=

# flag to ignore examples folder
ANOMSTACK_IGNORE_EXAMPLES=no

# email alerting related env vars
ANOMSTACK_ALERT_EMAIL_FROM=
ANOMSTACK_ALERT_EMAIL_TO=
ANOMSTACK_ALERT_EMAIL_SMTP_HOST=smtp.gmail.com
ANOMSTACK_ALERT_EMAIL_SMTP_PORT=587
ANOMSTACK_ALERT_EMAIL_PASSWORD=

# email failure alerting related env vars
# ANOMSTACK_FAILURE_EMAIL_FROM=
# ANOMSTACK_FAILURE_EMAIL_TO=
# ANOMSTACK_FAILURE_EMAIL_SMTP_HOST=smtp.gmail.com
# ANOMSTACK_FAILURE_EMAIL_SMTP_PORT=587
# ANOMSTACK_FAILURE_EMAIL_PASSWORD=

# dagster related env vars - LIGHTWEIGHT DEFAULTS
# Use smaller, more manageable directories to prevent disk space issues
ANOMSTACK_DAGSTER_LOCAL_ARTIFACT_STORAGE_DIR=tmp_light/artifacts
ANOMSTACK_DAGSTER_OVERALL_CONCURRENCY_LIMIT=5  # Reduced from 10
ANOMSTACK_DAGSTER_DEQUEUE_USE_THREADS=false
ANOMSTACK_DAGSTER_DEQUEUE_NUM_WORKERS=2  # Reduced from 4
ANOMSTACK_DAGSTER_LOCAL_COMPUTE_LOG_MANAGER_DIRECTORY=tmp_light/compute_logs
ANOMSTACK_DAGSTER_SQLITE_STORAGE_BASE_DIR=tmp_light/storage

# =============================================================================
# LLM Alert Configuration (anomaly-agent)
# =============================================================================

# LLM Platform: "openai" (default) or "openrouter"
ANOMSTACK_LLM_PLATFORM=openai

# API Key (works for both OpenAI and OpenRouter)
OPENAI_API_KEY=

# Model selection for LLM alerts
# Set to "random" to A/B test across multiple models
# Comma-separated list of models when using random selection
# With OpenRouter, use provider/model format:
#   openai/gpt-4o-mini,anthropic/claude-3.5-sonnet,google/gemini-2.0-flash-exp
ANOMSTACK_LLMALERT_MODELS=gpt-4o-mini

# LangSmith tracing (optional)
LANGSMITH_TRACING=true
LANGSMITH_ENDPOINT="https://api.smith.langchain.com"
LANGSMITH_API_KEY=""
LANGSMITH_PROJECT="anomaly-agent"

# Legacy config (deprecated)
ANOMSTACK_OPENAI_KEY=
ANOMSTACK_OPENAI_MODEL=gpt-4o-mini
ANOMSTACK_ANTHROPIC_KEY=
ANOMSTACK_ANTHROPIC_MODEL=claude-3-haiku-20240307

# some dagster env vars
DAGSTER_LOG_LEVEL=DEBUG
DAGSTER_CONCURRENCY=4
# Code server host for workspace configuration
# For Docker Compose: leave blank (defaults to "anomstack_code")
# For Fly.io single container: set to "localhost"
DAGSTER_CODE_SERVER_HOST=

# max runtime for a job in dagster
# https://docs.dagster.io/deployment/run-monitoring#general-run-timeouts
ANOMSTACK_MAX_RUNTIME_SECONDS_TAG=900
# kill runs that exceed this many minutes
ANOMSTACK_KILL_RUN_AFTER_MINUTES=15

# ===== RETENTION CONFIGURATION =====
# Custom retention job configuration (SQLite doesn't support built-in Dagster retention)
# Number of days to keep Dagster runs (default: 1)
ANOMSTACK_RETENTION_DAYS=1
# Cron schedule for retention job (default: every 6 hours)
ANOMSTACK_RETENTION_SCHEDULE=0 */6 * * *
# Whether retention is enabled (default: true)
ANOMSTACK_RETENTION_ENABLED=true
# Whether to vacuum SQLite database after deletions (default: true)
ANOMSTACK_RETENTION_VACUUM=true
# Batch size for deletions to manage memory usage (default: 1000)
ANOMSTACK_RETENTION_BATCH_SIZE=1000
# Base path to Dagster storage (default: /data/dagster_storage)
ANOMSTACK_DAGSTER_STORAGE_PATH=/data/dagster_storage

# postgres related env vars
ANOMSTACK_POSTGRES_USER=postgres_user
ANOMSTACK_POSTGRES_PASSWORD=postgres_password
ANOMSTACK_POSTGRES_DB=postgres_db
# leave blank if you don't want to forward the port locally for monitoring etc.
ANOMSTACK_POSTGRES_FORWARD_PORT=5432

# motherduck related env vars
ANOMSTACK_MOTHERDUCK_TOKEN=

# turso related env vars
ANOMSTACK_TURSO_DATABASE_URL=
ANOMSTACK_TURSO_AUTH_TOKEN=

# clickhouse related env vars
ANOMSTACK_CLICKHOUSE_HOST=localhost
ANOMSTACK_CLICKHOUSE_PORT=8123
ANOMSTACK_CLICKHOUSE_USER=anomstack
ANOMSTACK_CLICKHOUSE_PASSWORD=anomstack
ANOMSTACK_CLICKHOUSE_DATABASE=default

# dashbooard port
ANOMSTACK_DASHBOARD_PORT = 5001

# PostHog API key for metrics ingestion (querying PostHog's API for data)
# Also used by anomaly-agent for LLM analytics tracking when POSTHOG_ENABLED=true
POSTHOG_API_KEY=

# PostHog frontend API key for analytics tracking (website analytics)
POSTHOG_FRONTEND_API_KEY=

# PostHog LLM Analytics for anomaly-agent (optional)
# Tracks LLM calls (tokens, latency, costs) via posthog.ai.langchain integration
POSTHOG_ENABLED=false
POSTHOG_HOST=https://us.i.posthog.com
POSTHOG_DISTINCT_ID=
# POSTHOG_PRIVACY_MODE=false  # Set true to not log prompt/response content

# example to enable some jobs via env vars
# ANOMSTACK__PYTHON_INGEST_SIMPLE__INGEST_DEFAULT_SCHEDULE_STATUS=RUNNING
# ANOMSTACK__PYTHON_INGEST_SIMPLE__TRAIN_DEFAULT_SCHEDULE_STATUS=RUNNING
# ANOMSTACK__PYTHON_INGEST_SIMPLE__SCORE_DEFAULT_SCHEDULE_STATUS=RUNNING
# ANOMSTACK__PYTHON_INGEST_SIMPLE__ALERT_DEFAULT_SCHEDULE_STATUS=RUNNING
# ANOMSTACK__PYTHON_INGEST_SIMPLE__PLOT_DEFAULT_SCHEDULE_STATUS=RUNNING
# ANOMSTACK__PYTHON_INGEST_SIMPLE__CHANGE_DEFAULT_SCHEDULE_STATUS=RUNNING
# ANOMSTACK__PYTHON_INGEST_SIMPLE__LLMALERT_DEFAULT_SCHEDULE_STATUS=RUNNING
# ANOMSTACK__PYTHON_INGEST_SIMPLE__DELETE_DEFAULT_SCHEDULE_STATUS=RUNNING

# automatic configuration reloading
# Enable scheduled config reload job (runs every N minutes)
ANOMSTACK_AUTO_CONFIG_RELOAD=false
ANOMSTACK_CONFIG_RELOAD_CRON="*/5 * * * *"
ANOMSTACK_CONFIG_RELOAD_STATUS=STOPPED

# Enable smart config file watcher sensor (reloads only when files change)
ANOMSTACK_CONFIG_WATCHER=true
ANOMSTACK_CONFIG_WATCHER_INTERVAL=30