Add Sentry error reporting documentation

src/tribler/core/sentry_reporter/README.md

@@ -0,0 +1,57 @@
+# Tribler Error Reporting with Sentry
+
+For collecting errors from users' machines, we use [Sentry](https://docs.sentry.io/). 
+Due to the specific nature of our application, we cannot automatically collect 
+errors (which is the default behavior of Sentry). Therefore, we had to modify 
+the usual Sentry mechanism.
+https://github.com/Tribler/tribler/blob/main/src/tribler/core/sentry_reporter/sentry_reporter.py
+
+The changes we made allow us to control Sentry and limit its error reporting
+capabilities. By default, in Tribler, sending Sentry messages is suppressed.
+When an error occurs, the user is presented with a dialog where they must explicitly
+indicate that the error should be sent to the Tribler team.
+
+## Modified Sentry Modes
+
+The modified Sentry has three operating modes:
+
+```python
+class SentryStrategy(Enum):
+    """Class describes all available Sentry Strategies
+
+    SentryReporter can work with 3 strategies:
+    1. Send reports are allowed
+    2. Send reports are allowed with a confirmation dialog
+    3. Send reports are suppressed (but the last event will be stored)
+    """
+
+    SEND_ALLOWED = auto()
+    SEND_ALLOWED_WITH_CONFIRMATION = auto()
+    SEND_SUPPRESSED = auto()
+```
+
+Immediately after Tribler starts, the mode is set to `SEND_ALLOWED_WITH_CONFIRMATION`. 
+As soon as the UI loads, Sentry switches to `SEND_SUPPRESSED`.
+
+`SEND_ALLOWED` is used only on developers' machines if they explicitly set 
+`TRIBLER_TEST_SENTRY_URL` in the environment variables.
+
+## Scrubber
+
+We use an advanced Scrubber to remove all sensitive parts from the user's report. 
+Our Scrubber processes all fields of the Sentry event and replaces sensitive parts 
+with placeholders.
+
+Parts considered sensitive:
+- User name
+- User machine name
+- User-specific folder names
+- IP addresses
+- Torrent hashes
+
+For placeholders, we use words generated by [Faker](https://pypi.org/project/Faker/) 
+based on the string. This way, we can understand when the same strings repeat in 
+the report, but we cannot determine what the actual value of the string was on 
+the user's machine.
+
+https://github.com/Tribler/tribler/blob/main/src/tribler/core/sentry_reporter/sentry_scrubber.py