Skip to content

[FEAT] Story 2.4 Scenario Context Testing & Refinement #21

@yeomin4242

Description

@yeomin4242

한줄 설명

Build comprehensive testing framework to validate AI prompt quality, scenario coherence, and character consistency across diverse scenarios with automated quality metrics.

문제·기회

No response

제안 내용

No response

완료 기준(AC)

  • ScenarioContextTester automated test suite with 30+ test scenarios
  • Test categories: Character consistency (10 tests), Event coherence (10 tests), Setting adaptation (10 tests)
  • Each test includes: scenario definition, expected AI behavior, evaluation criteria
  • Automated quality metrics: Coherence score (1-10), Character consistency score (1-10), Creativity score (1-10)
  • Gemini 2.5 Flash as judge: Meta-prompting to evaluate AI responses for quality
  • Test report generation: JSON output with pass/fail status, scores, example responses
  • Regression testing: Compare new prompt versions against baseline quality
  • /api/ai/test-scenario admin endpoint to run individual scenario tests
  • CI/CD integration: Run test suite on prompt template changes
  • Quality threshold: Average score ≥ 7.0 required to pass

관련 참고자료

No response

관련 이슈·블로커

No response

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions