Use LLMs (OpenAI, Google Gemini, or Anthropic Claude) to generate test cases from a user story. Support outputs in plain English (descriptive narratives) or BDD/Gherkin style (e.g., Feature/Scenario/Given/When/Then). Optionally integrate with Jira to fetch user stories, prerequisites, instructions, etc.