refactor: Improve AC scoping for autonomous LLM implementation

[shinpr]

Summary

Refine Acceptance Criteria (AC) scoping guidelines to focus on high-ROI test coverage in
LLM/CI/CD environments. This improves autonomous implementation accuracy by clarifying what
should and shouldn't be included in ACs.

Changes

acceptance-test-generator.md (EN/JA)

• Simplified "Out of Scope" section into 3 clear categories:
  • External Dependencies (test contracts/interfaces instead)
  • Non-Deterministic in CI (performance metrics, load testing)
  • Implementation Details (not user-observable)
• Removed redundant section: "Security and Performance Requirements Processing"
• Added Action guideline: How to handle excluded items in ACs

technical-designer.md (EN/JA)

• Added "AC Scoping for Autonomous Implementation" section:
  • Include (High automation ROI): Business logic, data integrity, user-visible
    functionality, error handling behavior
  • Exclude (Low ROI in LLM/CI/CD): External service connections, performance metrics,
    implementation details, UI layout specifics
  • Principle: AC = User-observable behavior verifiable in isolated CI environment

Version

• Bumped version from 1.7.13 to 1.7.14

Rationale

• Focus on observable behavior: ACs should specify what users see/experience, not how
  it's implemented
• Optimize for CI/CD: Exclude non-deterministic tests that fail in automated
  environments
• Improve automation ROI: Concentrate testing effort on business logic and data
  integrity
• Align with LLM capabilities: Match AC scope to what autonomous agents can effectively
  verify

Impact

• For AC authors: Clearer guidelines on what to include/exclude in acceptance criteria
• For test generators: Reduced ambiguity when transforming ACs into test cases
• For autonomous implementation: Better alignment between requirements and verifiable
  outcomes