feat: per-turn pass/fail display in multi-turn console output

[hidai25]

Summary

Right now, multi-turn tests show a single merged result in the console — but there's no visibility into which individual turn caused a failure.

Current behaviour

✗  flight-booking-conversation   score=52  (3 turns)

Desired behaviour

✗  flight-booking-conversation   score=52  (3 turns)
     Turn 1 ✓  search_flights called, output OK
     Turn 2 ✗  book_flight not called  ← failure here
     Turn 3 –  skipped (previous turn failed)

Implementation hints

• _execute_multi_turn_trace() in evalview/cli.py already collects turn_traces (one ExecutionTrace per turn)
• The per-turn traces are merged before evaluation — split them out first, evaluate each independently, then merge
• The verbose log line is at cli.py around the is_multi_turn dispatch block — add per-turn status lines there with [dim] formatting
• Use ✓ / ✗ / – glyphs, indented 5 spaces to align under the test name

Files to touch

• evalview/cli.py — execute_single_test + _execute_multi_turn_trace
• Optionally evalview/reporters/console_reporter.py if you want a shared formatter

Good first issue?

Yes — the plumbing (turn_traces list) is already there. This is purely a display change with no changes to evaluation logic.

Acceptance criteria

• evalview run shows per-turn status for multi-turn tests
• Pass/fail per turn is visible even without --verbose
• No changes to scoring logic