[Add] browser-use and main.py

browser-use/browser_use/telemetry/service.py

@@ -0,0 +1,124 @@
+import logging
+import os
+import uuid
+from pathlib import Path
+
+from dotenv import load_dotenv
+from posthog import Posthog
+
+from browser_use.telemetry.views import BaseTelemetryEvent
+from browser_use.utils import singleton
+
+load_dotenv()
+
+
+logger = logging.getLogger(__name__)
+
+
+POSTHOG_EVENT_SETTINGS = {
+	'process_person_profile': True,
+}
+
+
+def xdg_cache_home() -> Path:
+	default = Path.home() / '.cache'
+	env_var = os.getenv('XDG_CACHE_HOME')
+	if env_var and (path := Path(env_var)).is_absolute():
+		return path
+	return default
+
+
+@singleton
+class ProductTelemetry:
+	"""
+	Service for capturing anonymized telemetry data.
+
+	If the environment variable `ANONYMIZED_TELEMETRY=False`, anonymized telemetry will be disabled.
+	"""
+
+	USER_ID_PATH = str(xdg_cache_home() / 'browser_use' / 'telemetry_user_id')
+	PROJECT_API_KEY = 'phc_F8JMNjW1i2KbGUTaW1unnDdLSPCoyc52SGRU0JecaUh'
+	HOST = 'https://eu.i.posthog.com'
+	UNKNOWN_USER_ID = 'UNKNOWN'
+
+	_curr_user_id = None
+
+	def __init__(self) -> None:
+		telemetry_disabled = os.getenv('ANONYMIZED_TELEMETRY', 'true').lower() == 'false'
+		self.debug_logging = os.getenv('BROWSER_USE_LOGGING_LEVEL', 'info').lower() == 'debug'
+
+		if telemetry_disabled:
+			self._posthog_client = None
+		else:
+			logger.info(
+				'Anonymized telemetry enabled. See https://docs.browser-use.com/development/telemetry for more information.'
+			)
+			self._posthog_client = Posthog(
+				project_api_key=self.PROJECT_API_KEY,
+				host=self.HOST,
+				disable_geoip=False,
+				enable_exception_autocapture=True,
+			)
+
+			# Silence posthog's logging
+			if not self.debug_logging:
+				posthog_logger = logging.getLogger('posthog')
+				posthog_logger.disabled = True
+
+		if self._posthog_client is None:
+			logger.debug('Telemetry disabled')
+
+	def capture(self, event: BaseTelemetryEvent) -> None:
+		if self._posthog_client is None:
+			return
+
+		if self.debug_logging:
+			logger.debug(f'Telemetry event: {event.name} {event.properties}')
+		self._direct_capture(event)
+
+	def _direct_capture(self, event: BaseTelemetryEvent) -> None:
+		"""
+		Should not be thread blocking because posthog magically handles it
+		"""
+		if self._posthog_client is None:
+			return
+
+		try:
+			self._posthog_client.capture(
+				self.user_id,
+				event.name,
+				{**event.properties, **POSTHOG_EVENT_SETTINGS},
+			)
+		except Exception as e:
+			logger.error(f'Failed to send telemetry event {event.name}: {e}')
+
+	def flush(self) -> None:
+		if self._posthog_client:
+			try:
+				self._posthog_client.flush()
+				logger.debug('PostHog client telemetry queue flushed.')
+			except Exception as e:
+				logger.error(f'Failed to flush PostHog client: {e}')
+		else:
+			logger.debug('PostHog client not available, skipping flush.')
+
+	@property
+	def user_id(self) -> str:
+		if self._curr_user_id:
+			return self._curr_user_id
+
+		# File access may fail due to permissions or other reasons. We don't want to
+		# crash so we catch all exceptions.
+		try:
+			if not os.path.exists(self.USER_ID_PATH):
+				os.makedirs(os.path.dirname(self.USER_ID_PATH), exist_ok=True)
+				with open(self.USER_ID_PATH, 'w') as f:
+					new_user_id = str(uuid.uuid4())
+					f.write(new_user_id)
+				self._curr_user_id = new_user_id
+			else:
+				with open(self.USER_ID_PATH) as f:
+					self._curr_user_id = f.read()
+		except Exception:
+			self._curr_user_id = 'UNKNOWN_USER_ID'
+		return self._curr_user_id

browser-use/browser_use/telemetry/views.py

@@ -0,0 +1,56 @@
+from abc import ABC, abstractmethod
+from collections.abc import Sequence
+from dataclasses import asdict, dataclass
+from typing import Any
+
+
+@dataclass
+class BaseTelemetryEvent(ABC):
+	@property
+	@abstractmethod
+	def name(self) -> str:
+		pass
+
+	@property
+	def properties(self) -> dict[str, Any]:
+		return {k: v for k, v in asdict(self).items() if k != 'name'}
+
+
+@dataclass
+class RegisteredFunction:
+	name: str
+	params: dict[str, Any]
+
+
+@dataclass
+class ControllerRegisteredFunctionsTelemetryEvent(BaseTelemetryEvent):
+	registered_functions: list[RegisteredFunction]
+	name: str = 'controller_registered_functions'
+
+
+@dataclass
+class AgentTelemetryEvent(BaseTelemetryEvent):
+	# start details
+	task: str
+	model: str
+	model_provider: str
+	planner_llm: str | None
+	max_steps: int
+	max_actions_per_step: int
+	use_vision: bool
+	use_validation: bool
+	version: str
+	source: str
+	# step details
+	action_errors: Sequence[str | None]
+	action_history: Sequence[list[dict] | None]
+	urls_visited: Sequence[str | None]
+	# end details
+	steps: int
+	total_input_tokens: int
+	total_duration_seconds: float
+	success: bool | None
+	final_result_response: str | None
+	error_message: str | None
+
+	name: str = 'agent_event'