v0.6.3

Added

• STEALTH_ENABLED — global stealth mode
  New Scrapy setting that routes all requests through the stealth engine automatically — no need
  to add meta={"stealth": {...}} on every request. Set once in settings.py or custom_settings:

  STEALTH_ENABLED = True

  Per-request opt-out is still supported via meta={"stealth": False}.
  STEALTH_ENABLED is also read from spider_opened so custom_settings on the spider class
  takes effect without restarting the crawler process.

• STEALTH_DRIVER — engine driver configurable from Scrapy settings
  The default stealth driver ("basic", "turbo", or "browser") can now be set globally in
  settings.py or custom_settings instead of per-request:

  STEALTH_DRIVER = "turbo"

  spider_opened reads this setting and applies it to config.STEALTH_DRIVER at spider start.

• StealthResponse — engine name in response flags
  Every StealthResponse now includes the engine driver name in its Scrapy flags list alongside
  the package logger name (e.g. ["scrapy-stealth", "turbo"]). The flag appears in Scrapy's
  crawl log next to each response line, making it easy to see which engine handled each request.

Fixed

• EngineManager.get() — KeyError crash on invalid driver
  When STEALTH_DRIVER was set to a typo (e.g. "browsesr") the fallback path also read
  config.get("STEALTH_DRIVER"), which returned the same invalid value, causing a KeyError.
  The fallback now uses _DEFAULT_DRIVER from constants (the package-level safe default),
  with config.STEALTH_DRIVER preferred when it is itself valid.