.env.example

# Model provider configuration
API_KEY=
API_BASE=http://mockserver:8090/v1

# Model configuration
# MODEL_PROVIDER selects the LLM integration (via LangChain init_chat_model).
# Built-in providers: openai, deepseek, anthropic, ollama. OpenAI-compatible
# endpoints (DeepSeek / OneAPI / vLLM / ...) work with openai + API_BASE.
# See docs/configuration.md for per-provider examples.
MODEL_PROVIDER=openai
MODEL_NAME=deepseek-chat
TEMPERATURE=0.7
MAX_TOKENS=2000

# LLM gateway provider: langchain (default) or openai.
# - langchain: uses init_chat_model, supports many providers via MODEL_PROVIDER.
# - openai:    talks to OpenAI / OpenAI-compatible endpoints (API_BASE) directly
#              via the official openai Python SDK (MODEL_PROVIDER is ignored).
#LLM_PROVIDER=langchain

# MongoDB configuration
#MONGODB_URI=mongodb://mongodb:27017
#MONGODB_DATABASE=manus
#MONGODB_USERNAME=
#MONGODB_PASSWORD=

# Redis configuration
#REDIS_HOST=redis
#REDIS_PORT=6379
#REDIS_DB=0
#REDIS_PASSWORD=

# Sandbox configuration
#SANDBOX_ADDRESS=
SANDBOX_IMAGE=simpleyyt/manus-sandbox
SANDBOX_NAME_PREFIX=sandbox
SANDBOX_TTL_MINUTES=30
SANDBOX_NETWORK=manus-network
#SANDBOX_CHROME_ARGS=
#SANDBOX_HTTPS_PROXY=
#SANDBOX_HTTP_PROXY=
#SANDBOX_NO_PROXY=

# Browser engine configuration
# Options: browser_use (default), playwright
# - browser_use: drives Chrome through the browser-use library's event bus
#                (official DOM serializer, watchdog-managed waits/tab switching)
# - playwright:  lightweight Playwright-over-CDP engine emitting the same
#                tool payload shape (element tree, Markdown content, console)
#BROWSER_ENGINE=browser_use

# Search engine configuration
# Options: baidu, baidu_web, google, bing, bing_web, tavily, serper, custom
# baidu:    uses the Baidu Qianfan AI Search API (requires BAIDU_SEARCH_API_KEY)
# baidu_web: scrapes Baidu search results with browser impersonation (no API key needed)
# bing:     uses the official Bing Web Search API (requires BING_SEARCH_API_KEY)
# bing_web: scrapes Bing search results directly (no API key needed)
# tavily:   uses the Tavily Search API (requires TAVILY_API_KEY)
# serper:   uses the Serper.dev Google Search API (requires SERPER_API_KEY)
# custom:   calls any third-party search REST API via SEARCH_API_URL + SEARCH_API_KEY
SEARCH_PROVIDER=bing_web

# Baidu search configuration, only used when SEARCH_PROVIDER=baidu
# Get your API key from https://console.bce.baidu.com/qianfan/ais/console/onlineService
#BAIDU_SEARCH_API_KEY=

# Bing search configuration, only used when SEARCH_PROVIDER=bing
# Get your API key from https://www.microsoft.com/en-us/bing/apis/bing-web-search-api
#BING_SEARCH_API_KEY=

# Google search configuration, only used when SEARCH_PROVIDER=google
#GOOGLE_SEARCH_API_KEY=
#GOOGLE_SEARCH_ENGINE_ID=

# Tavily search configuration, only used when SEARCH_PROVIDER=tavily
# Get your API key from https://tavily.com
#TAVILY_API_KEY=

# Serper.dev search configuration, only used when SEARCH_PROVIDER=serper
# Returns reliable Google results. Get your API key from https://serper.dev
#SERPER_API_KEY=

# Custom search API configuration, only used when SEARCH_PROVIDER=custom
# Allows integration with any third-party search REST API.
#
# Minimal setup (POST + Bearer token, e.g. a custom internal API):
#   SEARCH_API_URL=https://your-search-api.example.com/search
#   SEARCH_API_KEY=your-api-key
#
# Serper.dev via custom provider:
#   SEARCH_API_URL=https://google.serper.dev/search
#   SEARCH_API_KEY=your-serper-key
#   SEARCH_API_KEY_HEADER=X-API-KEY
#   SEARCH_API_KEY_HEADER_PREFIX=
#   SEARCH_RESULT_FIELD=organic
#
# SerpAPI via custom provider:
#   SEARCH_API_URL=https://serpapi.com/search
#   SEARCH_API_KEY=your-serpapi-key
#   SEARCH_API_KEY_PARAM=api_key
#   SEARCH_API_METHOD=GET
#   SEARCH_RESULT_FIELD=organic_results
#
# Brave Search API via custom provider:
#   SEARCH_API_URL=https://api.search.brave.com/res/v1/web/search
#   SEARCH_API_KEY=your-brave-key
#   SEARCH_API_KEY_HEADER=X-Subscription-Token
#   SEARCH_API_KEY_HEADER_PREFIX=
#   SEARCH_API_METHOD=GET
#   SEARCH_RESULT_FIELD=web.results
#   SEARCH_SNIPPET_FIELD=description
#
#SEARCH_API_URL=
#SEARCH_API_KEY=
#SEARCH_API_KEY_HEADER=Authorization
#SEARCH_API_KEY_HEADER_PREFIX=Bearer
#SEARCH_API_KEY_PARAM=
#SEARCH_API_METHOD=POST
#SEARCH_QUERY_FIELD=q
#SEARCH_RESULT_FIELD=results
#SEARCH_TITLE_FIELD=title
#SEARCH_LINK_FIELD=link
#SEARCH_SNIPPET_FIELD=snippet

# Google Analytics configuration
# Set your Google Analytics Measurement ID (e.g. G-XXXXXXXXXX)
#GOOGLE_ANALYTICS_ID=

# Auth configuration
# Options: password, none, local
AUTH_PROVIDER=password

# Password auth configuration, only used when AUTH_PROVIDER=password
PASSWORD_SALT=
PASSWORD_HASH_ROUNDS=10

# Local auth configuration, only used when AUTH_PROVIDER=local
#LOCAL_AUTH_EMAIL=admin@example.com
#LOCAL_AUTH_PASSWORD=admin

# JWT configuration (URL signing + optional login grace for legacy tokens)
JWT_SECRET_KEY=your-secret-key-here
JWT_ALGORITHM=HS256
JWT_ACCESS_TOKEN_EXPIRE_MINUTES=30
JWT_REFRESH_TOKEN_EXPIRE_DAYS=7

# Opaque Redis auth sessions (browser HttpOnly Cookie + App Bearer)
# /auth/login returns access_token = opaque session_id (not a JWT).
# Browser: Cookie session_id; App: Authorization: Bearer <session_id>
# WS: same Cookie / Bearer header — no ?token= query
#SESSION_COOKIE_NAME=session_id
#SESSION_WEB_TTL_DAYS=14
#SESSION_APP_TTL_DAYS=30
#SESSION_COOKIE_SECURE=false
#SESSION_COOKIE_SAMESITE=lax
#SESSION_JWT_GRACE_ENABLED=true

# Email configuration
# Only used when AUTH_PROVIDER=password
#EMAIL_HOST=smtp.gmail.com
#EMAIL_PORT=587
#EMAIL_USERNAME=your-email@gmail.com
#EMAIL_PASSWORD=your-password
#EMAIL_FROM=your-email@gmail.com

# Extra headers for LLM API requests (JSON format)
#EXTRA_HEADERS={"X-Custom-Header": "value"}

# Task backend configuration
# local: run agent tasks in-process (default)
# celery: run agent tasks on distributed Celery workers
#         (requires a worker container, see docs/configuration.md)
#TASK_BACKEND=local
# Optional custom Celery broker URL (defaults to the Redis settings above)
#CELERY_BROKER_URL=

# MCP configuration
#MCP_CONFIG_PATH=/etc/mcp.json

# Log configuration
LOG_LEVEL=INFO