Merge pull request #5 from epam/feature/arbitrary_testspec_support

Make possible evaluation against arbitrary test scenario specification

correction of the singular form of the word criterion

feat: expose more types to give flexibility to the client

feat (grading): remove LLM grading to reduce the variability of the results (#3)

Merge pull request #2 from epam/feat/update-evaluation-prompt

feat (evaluation): adjust evaluation prompt

Add EvaluationReport class for parsing unstrctured output

Initial commit