fix: improve repair feedback for LLM-as-a-Judge validations

[akihikokuroda]

Pull Request

Issue

Fix: #1247

Description

When LLM-as-a-Judge validation returns binary yes/no answers, use the requirement
description as the repair reason instead of just 'yes' or 'no'. This applies to all
semantic validations (check(), req() without validation_fn, etc.) regardless of
check_only setting.

Previously, repair feedback would show:

     The following requirements failed before:
     * no

Now it shows:

     The following requirements failed before:
     * The email should have a salutation

This makes repair feedback more actionable and helps models understand what needs
to be fixed during repair iterations in RepairTemplateStrategy and MultiTurnStrategy.

Testing

• Tests added to the respective file if code was changed
• New code has 100% coverage if code was added
• Ensure existing tests and github automation passes (a maintainer will kick off the github automation when the rest of the PR is populated)

Attribution

• AI coding assistants used: claudecode

---

Adding a new component, requirement, sampling strategy, or tool?

If your PR adds or modifies one of the types below, check the matching box. A checklist of type-specific review items will be posted as a comment.

• Component
• Requirement
• Sampling Strategy
• Tool

NOTE: Please ensure you have an issue that has been acknowledged by a core contributor and routed you to open a pull request against this repository. Otherwise, please open an issue before continuing with this pull request.