flutter_test_gen_ai 0.2.4

An LLM-based test generation tool that generates Dart test cases for uncovered code using Google Gemini to improve code coverage.

0.2.4

• Improved subprocess diagnostics by reporting standard output (stdout) alongside standard error (stderr) when flutter test --coverage fails.
• Handled ProcessException gracefully at the CLI entry point to print user-friendly troubleshooting details instead of crashing with a raw unhandled exception.

0.2.3

• Registered public library exports in lib/flutter_test_gen_ai.dart to support programmatic scripting and library integration.
• Added a compile-safe programmatic usage demonstration to example/example.dart.
• Added an explanation section at the bottom of the generated markdown reports (testgen_report.md) detailing why some tests fail (e.g., hardcoded global standard I/O, platform channels).
• Cleaned up package footprint using .pubignore to exclude local model assets and development configs.
• Disables global CLI configuration in pubspec.yaml to ensure pub.dev displays standard dev_dependencies installation instructions.

0.2.2

• Fixed analyzer API compatibility issues for newer Dart/Flutter SDKs.
• Focused documentation and usage instructions purely on local CLI execution.

0.2.0

• Added robust support for automated Flutter Widget Testing.
• Static analyzer dynamically extracts Semantics labels and widget Keys from source code to guide Gemini's test generation.
• Prevents selector hallucinations in generated widget tests.
• Fixed UI hierarchy resolution for nested widget methods.

0.1.0

• Initial release: LLM-based coverage-driven test generation CLI tool.
• Supports both Flutter and pure Dart packages.
• Added AST-driven static dependency context construction.
• Automated self-correcting validation loop.
• Optional visual Mermaid dependency and coverage reporting.