feat: add Docker configuration files and installation script for SurfSense

docker/.env.example

@@ -0,0 +1,194 @@
+# ==============================================================================
+# SurfSense Docker Configuration
+# ==============================================================================
+# Only variables YOU need to set are in this file.
+# Database, Redis, and internal service wiring are handled automatically.
+# ==============================================================================
+
+# SurfSense version (pin to a specific version like "0.0.13.1" or use "latest")
+SURFSENSE_VERSION=latest
+
+# ------------------------------------------------------------------------------
+# Core Settings
+# ------------------------------------------------------------------------------
+
+# REQUIRED: Generate a secret key with: openssl rand -base64 32
+SECRET_KEY=replace_me_with_a_random_string
+
+# Auth type: LOCAL (email/password) or GOOGLE (OAuth)
+AUTH_TYPE=LOCAL
+
+# Allow new user registrations (TRUE or FALSE)
+# REGISTRATION_ENABLED=TRUE
+
+# Document parsing service: DOCLING, UNSTRUCTURED, or LLAMACLOUD
+ETL_SERVICE=DOCLING
+
+# Embedding model for vector search
+# Local: sentence-transformers/all-MiniLM-L6-v2
+# OpenAI: openai://text-embedding-ada-002 (set OPENAI_API_KEY below)
+# Cohere: cohere://embed-english-light-v3.0 (set COHERE_API_KEY below)
+EMBEDDING_MODEL=sentence-transformers/all-MiniLM-L6-v2
+
+# ------------------------------------------------------------------------------
+# Ports (change these to avoid host conflicts — everything auto-derives)
+# ------------------------------------------------------------------------------
+
+# BACKEND_PORT=8000
+# FRONTEND_PORT=3000
+# ELECTRIC_PORT=5133
+
+# Frontend URL used by backend for CORS and OAuth redirects.
+# Auto-derived from FRONTEND_PORT for localhost. Set explicitly for reverse proxy.
+# NEXT_FRONTEND_URL=http://localhost:3000
+
+# Backend URL for OAuth callback redirects (set when behind a reverse proxy)
+# BACKEND_URL=https://api.yourdomain.com
+
+# ------------------------------------------------------------------------------
+# Frontend URL Overrides (reverse proxy / custom domains)
+# ------------------------------------------------------------------------------
+# These are auto-derived from the port settings above for localhost deployments.
+# You only need to set these explicitly when using a reverse proxy with real
+# domains (e.g. Caddy, Nginx, Cloudflare Tunnel).
+#
+# NEXT_PUBLIC_FASTAPI_BACKEND_URL=https://api.yourdomain.com
+# NEXT_PUBLIC_ELECTRIC_URL=https://electric.yourdomain.com
+
+# ------------------------------------------------------------------------------
+# Database (defaults work out of the box, change for security)
+# ------------------------------------------------------------------------------
+
+# DB_PASSWORD=surfsense
+
+# ------------------------------------------------------------------------------
+# TTS & STT (Text-to-Speech / Speech-to-Text)
+# ------------------------------------------------------------------------------
+
+# Local Kokoro TTS (default) or LiteLLM provider
+TTS_SERVICE=local/kokoro
+# TTS_SERVICE_API_KEY=
+# TTS_SERVICE_API_BASE=
+
+# Local Faster-Whisper STT: local/MODEL_SIZE (tiny, base, small, medium, large-v3)
+STT_SERVICE=local/base
+# Or use LiteLLM: openai/whisper-1
+# STT_SERVICE_API_KEY=
+# STT_SERVICE_API_BASE=
+
+# ------------------------------------------------------------------------------
+# Rerankers (optional, disabled by default)
+# ------------------------------------------------------------------------------
+
+# RERANKERS_ENABLED=TRUE
+# RERANKERS_MODEL_NAME=ms-marco-MiniLM-L-12-v2
+# RERANKERS_MODEL_TYPE=flashrank
+
+# ------------------------------------------------------------------------------
+# Google OAuth (only if AUTH_TYPE=GOOGLE)
+# ------------------------------------------------------------------------------
+
+# GOOGLE_OAUTH_CLIENT_ID=
+# GOOGLE_OAUTH_CLIENT_SECRET=
+
+# ------------------------------------------------------------------------------
+# Connector OAuth Keys (uncomment connectors you want to use)
+# ------------------------------------------------------------------------------
+
+# -- Google Connectors --
+# GOOGLE_CALENDAR_REDIRECT_URI=http://localhost:8000/api/v1/auth/google/calendar/connector/callback
+# GOOGLE_GMAIL_REDIRECT_URI=http://localhost:8000/api/v1/auth/google/gmail/connector/callback
+# GOOGLE_DRIVE_REDIRECT_URI=http://localhost:8000/api/v1/auth/google/drive/connector/callback
+
+# -- Notion --
+# NOTION_CLIENT_ID=
+# NOTION_CLIENT_SECRET=
+# NOTION_REDIRECT_URI=http://localhost:8000/api/v1/auth/notion/connector/callback
+
+# -- Slack --
+# SLACK_CLIENT_ID=
+# SLACK_CLIENT_SECRET=
+# SLACK_REDIRECT_URI=http://localhost:8000/api/v1/auth/slack/connector/callback
+
+# -- Discord --
+# DISCORD_CLIENT_ID=
+# DISCORD_CLIENT_SECRET=
+# DISCORD_REDIRECT_URI=http://localhost:8000/api/v1/auth/discord/connector/callback
+# DISCORD_BOT_TOKEN=
+
+# -- Atlassian (Jira & Confluence) --
+# ATLASSIAN_CLIENT_ID=
+# ATLASSIAN_CLIENT_SECRET=
+# JIRA_REDIRECT_URI=http://localhost:8000/api/v1/auth/jira/connector/callback
+# CONFLUENCE_REDIRECT_URI=http://localhost:8000/api/v1/auth/confluence/connector/callback
+
+# -- Linear --
+# LINEAR_CLIENT_ID=
+# LINEAR_CLIENT_SECRET=
+# LINEAR_REDIRECT_URI=http://localhost:8000/api/v1/auth/linear/connector/callback
+
+# -- ClickUp --
+# CLICKUP_CLIENT_ID=
+# CLICKUP_CLIENT_SECRET=
+# CLICKUP_REDIRECT_URI=http://localhost:8000/api/v1/auth/clickup/connector/callback
+
+# -- Airtable --
+# AIRTABLE_CLIENT_ID=
+# AIRTABLE_CLIENT_SECRET=
+# AIRTABLE_REDIRECT_URI=http://localhost:8000/api/v1/auth/airtable/connector/callback
+
+# -- Microsoft Teams --
+# TEAMS_CLIENT_ID=
+# TEAMS_CLIENT_SECRET=
+# TEAMS_REDIRECT_URI=http://localhost:8000/api/v1/auth/teams/connector/callback
+
+# -- Composio --
+# COMPOSIO_API_KEY=
+# COMPOSIO_ENABLED=TRUE
+# COMPOSIO_REDIRECT_URI=http://localhost:8000/api/v1/auth/composio/connector/callback
+
+# ------------------------------------------------------------------------------
+# External API Keys (optional)
+# ------------------------------------------------------------------------------
+
+# Firecrawl (web scraping)
+# FIRECRAWL_API_KEY=
+
+# Unstructured (if ETL_SERVICE=UNSTRUCTURED)
+# UNSTRUCTURED_API_KEY=
+
+# LlamaCloud (if ETL_SERVICE=LLAMACLOUD)
+# LLAMA_CLOUD_API_KEY=
+
+# ------------------------------------------------------------------------------
+# Observability (optional)
+# ------------------------------------------------------------------------------
+
+# LANGSMITH_TRACING=true
+# LANGSMITH_ENDPOINT=https://api.smith.langchain.com
+# LANGSMITH_API_KEY=
+# LANGSMITH_PROJECT=surfsense
+
+# ------------------------------------------------------------------------------
+# Advanced (optional)
+# ------------------------------------------------------------------------------
+
+# Periodic connector sync interval (default: 5m)
+# SCHEDULE_CHECKER_INTERVAL=5m
+
+# JWT token lifetimes
+# ACCESS_TOKEN_LIFETIME_SECONDS=86400
+# REFRESH_TOKEN_LIFETIME_SECONDS=1209600
+
+# Pages limit per user for ETL (default: unlimited)
+# PAGES_LIMIT=500
+
+# Connector indexing lock TTL in seconds (default: 28800 = 8 hours)
+# CONNECTOR_INDEXING_LOCK_TTL_SECONDS=28800
+
+# Residential proxy for web crawling
+# RESIDENTIAL_PROXY_USERNAME=
+# RESIDENTIAL_PROXY_PASSWORD=
+# RESIDENTIAL_PROXY_HOSTNAME=
+# RESIDENTIAL_PROXY_LOCATION=
+# RESIDENTIAL_PROXY_TYPE=1