[axxonite]

Describe what the PR does:

Follow-up to #792, as offered there. Adds a battery_replaced_timestamp property to
Tile, plus its as_dict() entry, README.md documentation and tests.

The Tile API records when a Tile's battery was last replaced, in
result.metadata.battery_replaced_at, as epoch milliseconds in a string. The property
returns it the same way last_timestamp and lost_timestamp are returned: a naive UTC
datetime, or None when the key is absent. Only Tiles with a user-replaceable battery
that has actually been replaced carry the key, so absence is the common case, not an
error.

Observed on a live account:

Tile	metadata
Keys	{"battery_state": "50", "battery_replaced_at": "1730646298403"}
Wallet	{"battery_state": "40", "battery_replaced_at": "1717167972131"}
Three sealed Tiles	{"battery_state": "10"}
Phone	{}

Tests: tile_details_response.json gains a battery_replaced_at value, so the
populated path is covered. tile_details_missing_last_state_response.json already has an
empty metadata block, which covers the None path.

Deliberately not included: metadata.battery_state. As described in #792, it is a
modelled estimate, not a measurement. A Tile that had not reported for 406 days, with a
flat cell, still showed "40". Tiles with no battery data show a placeholder "10", the
same value this repo's fixture carries next to "battery_status": "NONE". A consumer
would reasonably read it as a battery percentage, and it would be wrong in exactly the
case that matters. battery_replaced_at is a recorded fact, so it is the one worth
exposing.

Does this fix a specific issue?

Relates to #107 (closed by #792). This adds the
second piece of battery information discussed there.

Checklist:

• Confirm that one or more new tests are written for the new functionality.
• Run tests and ensure everything passes (with 100% test coverage).
• Update README.md with any new documentation.

Verified locally by running the repo's full pre-commit suite through Poetry against
dev (every hook passes, including pylint), and pytest (12 passed, 100% coverage).

🤖 Generated with Claude Code

https://claude.ai/code/session_01Qui3EQaoxiVRCBVVe9waLw

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 100.00%. Comparing base (ba4ed44) to head (28c92e2).

Additional details and impacted files

@@            Coverage Diff            @@
##               dev      #793   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files            5         5           
  Lines          181       187    +6     
=========================================
+ Hits           181       187    +6     

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.