DI: report evaluation errors in the correct schema

[p-datadog]

What does this PR do?
Changes the format of evaluation errors in the payload to match the format used by other languages (I used python for reference)

Motivation:
Implementation correctness

Change log entry
None (expression language evaluation has not yet shipped to customers)

Additional Notes:
The format in master was simple text messages I used to get expression language implementation off the ground.

How to test the change?

Tests added

[github-actions]

Thank you for updating Change log entry section 👏

^{Visited at: 2025-10-06 20:38:10 UTC}

[datadog-official]

✅ Tests

🎉 All green!

❄️ No new flaky tests detected
🧪 All tests passed

🎯 Code Coverage
• Patch Coverage: 103.03%
• Total Coverage: 101.96% (-0.00%)

View detailed report

_{This comment will be updated automatically if new data arrives.
🔗 Commit SHA: eb7c8c5 | Docs | Was this helpful? Give us feedback!}

[pr-commenter]

Benchmarks

Benchmark execution time: 2025-10-06 18:46:02

Comparing candidate commit eb7c8c5 in PR branch di-eval-errors-schema with baseline commit 7cce081 in branch master.

Found 0 performance improvements and 0 performance regressions! Performance is the same for 44 metrics, 2 unstable metrics.