fix: persist telemetry state as utf-8

[Ghraven]

What changed

• Read and write the server telemetry state file with explicit encoding="utf-8".
• Added a focused test that round-trips non-ASCII telemetry state through _save_state() and _load_state().

Problem

The telemetry state file stores JSON text but currently relies on the platform default encoding. On non-UTF-8 systems, state containing non-ASCII operator input, such as onboarding use-case text, can be decoded or written inconsistently.

Before/after

Before: STATE_PATH.read_text() and STATE_PATH.write_text(...) used the process default encoding.
After: telemetry state JSON is consistently read and written as UTF-8.

Verification

• python -m pytest tests/test_server_telemetry.py

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[Ghraven]

I have read the CLA Document and I hereby sign the CLA