Add support for thought signatures and thought tokens

[felixarntz]

This PR does two decoupled but related things:

1. Add support for thought signatures in message parts
   • This is technically more of a bug fix than an enhancement - it's a critical must-have. All of Anthropic, Google, and OpenAI rely heavily in thought signatures nowadays to validate that thought information passed back in a multi-turn chat was actually produced by their own APIs (rather than "made up" by the caller).
   • Contrary to what one might think, this is not only important for thought channel message parts. Several models will include thought signatures on actual content parts that (I suppose) involved thinking behind the scenes. For example, function calls commonly include them, but even images in multimodal output models can include them.
   • The rule is basically: If the API returns a thought signature with a message part, it must be passed when using the same content part in a response. Otherwise, the API call may result in an error.
2. Add support for thought tokens in token usage data

Both changes are fully optional and therefore backward compatible. However, once this is launched, we should update the provider implementations to support both where applicable - especially critical for point 1.

[github-actions]

The following accounts have interacted with this PR and/or linked issues. I will continue to update these lists as activity occurs. You can also manually ask me to refresh this list by adding the props-bot label.

If you're merging code through a pull request on GitHub, copy and paste the following into the bottom of the merge commit message.

Co-authored-by: felixarntz <flixos90@git.wordpress.org>
Co-authored-by: gziolo <gziolo@git.wordpress.org>

To understand the WordPress project's expectations around crediting contributors, please review the Contributor Attribution page in the Core Handbook.

[codecov]

Codecov Report

❌ Patch coverage is 89.18919% with 4 lines in your changes missing coverage. Please review.
✅ Project coverage is 87.61%. Comparing base (30a1506) to head (7352d45).
⚠️ Report is 19 commits behind head on trunk.

Files with missing lines	Patch %	Lines
src/Messages/DTO/MessagePart.php	81.81%	4 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff              @@
##              trunk     #216      +/-   ##
============================================
+ Coverage     87.45%   87.61%   +0.16%     
- Complexity     1194     1209      +15     
============================================
  Files            60       60              
  Lines          3858     3924      +66     
============================================
+ Hits           3374     3438      +64     
- Misses          484      486       +2     

Flag	Coverage Δ
unit	87.61% <89.18%> (+0.16%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 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.

[gziolo]

Looks good — clean, backward compatible, well-tested. All checks pass locally (tests, PHPCS clean, PHPStan clean). Left a minor nit on test values.

[gziolo]

Minor nit on tests/unit/Results/DTO/TokenUsageTest.php line 320 and a few other places:

$tokenUsage = new TokenUsage(100, 50, 250, 100);

thoughtTokens (100) exceeds completionTokens (50), which doesn't match the class docs stating thought tokens are "a subset of completion tokens." Might be worth adjusting for clarity — e.g. new TokenUsage(100, 150, 250, 100) — so the test values reinforce the documented invariant. Not blocking.