Expand evaluation cost reporting

[ShuxinLin]

Description

• add pricing for new Claude, GPT, Gemini, and MiniMax models
• split aggregate estimated cost into input and output totals
• add per-runner/model pass rates, operational metrics, cost per scenario, and cost per pass to score summaries
• render per-runner/model pass counts in the text summary
• add regression coverage for cost components and report fields

Type of Change

• New Benchmark Scenario (Industry/Asset type)
• Evaluation Metric / Scorer
• Agentic Orchestration Logic (ReAct, Plan-Execute, etc.)
• Infrastructure / Tooling Improvement

Industry Relevance

Makes benchmark quality and estimated operating cost directly comparable across runner/model combinations.

Related Issues

None.

Testing & Validation

• Unit Tests: uv run pytest src/evaluation/tests/test_metrics.py src/evaluation/tests/test_report.py (18 passed)
• Full evaluation suite: uv run pytest src/evaluation/tests (90 passed, 7 unrelated failures: one reflects the existing local lite.yaml edit; six exercise static-JSON CAR metadata not implemented on the base branch)
• Scenario Validation: Not run; this change does not alter agent trajectory execution.
• Data Integrity: No PII or sensitive industrial data is included.

Checklist

• My code follows the project formatting conventions and passes git diff --check.
• I have performed a self-review of my code.
• I have updated the documentation (README or /docs) accordingly.
• I have signed off my commits (DCO).