1.5.4 — Test runner configuration updates and repository cleanup (coverage docs removed)

Main changes

Updated Vitest configuration (more stable CI/local runs)

vitest.config.ts was tuned to reduce flakiness and resource issues:

• Runs tests in a fork pool (pool: 'forks') with constrained parallelism (maxForks: 2, minForks: 1) to avoid memory spikes.
• Adds explicit timeouts:
  • testTimeout: 30000
  • hookTimeout: 10000
  • teardownTimeout: 10000
• Sets TZ to America/New_York for consistent timezone-sensitive behavior.
• Keeps V8 coverage enabled and sets low baseline thresholds (20% across statements/branches/functions/lines), which may affect pipelines that enforce vitest thresholds.

Impact

• Developers should see fewer intermittent test failures and fewer “OOM”-style issues when running tests with coverage.
• CI timing characteristics may change slightly due to constrained parallelism and explicit timeouts.

Package metadata / dependency updates

package.json was updated with dependency/metadata adjustments (no code-level API changes were part of this release based on the diff). The published version is 1.5.4.

Documentation / repo hygiene

Removed generated coverage report documents from the repository

Several large, generated coverage report/plan markdown files were removed:

• TEST_COVERAGE_SUMMARY.md
• FINAL_COVERAGE_REPORT.md
• COVERAGE_IMPROVEMENT.md
• multiple docs/coverage-reflection-*.md artifacts

Impact

• If you referenced these files directly (links from READMEs, internal docs, or tooling expecting them in the repo), those references will need to be updated.
• Coverage reporting itself is still supported via Vitest; this change is about removing committed report artifacts.

Breaking changes

No breaking changes were detected in the release range (no production source/API changes were modified in this diff). The primary behavioral changes are in test execution configuration and the removal of documentation artifacts.