test: add 6× ultra-premium multiplier tests (#1186)

[microsasa]

Closes #1186

Changes

Adds two missing tests that exercise the claude-opus-4.6-1m model's 6× multiplier through the cost-estimation pipeline:

1. TestEstimatePremiumCost::test_ultra_premium_tier_model_uses_6x_multiplier — verifies _estimate_premium_cost correctly applies the 6.0× multiplier (not capped at 3×) for multiple call counts.

2. TestRenderLiveSessions::test_est_cost_ultra_premium_model — integration test confirming render_live_sessions displays ~30 (5 calls × 6.0) and does not show ~15 (which would indicate a 3× cap).

These tests guard against the regression scenario described in the issue where refactoring _estimate_premium_cost to use pricing.tier instead of pricing.multiplier would silently underestimate costs for ultra-premium models.

Warning

⚠️ Firewall blocked 2 domains

The following domains were blocked by the firewall during workflow execution:

• astral.sh
• pypi.org

To allow these domains, add them to the network.allowed list in your workflow frontmatter:

network:
  allowed:
    - defaults
    - "astral.sh"
    - "pypi.org"

See Network Configuration for more information.

Generated by Issue Implementer · ● 6M · ◷

[Copilot]

Pull request overview

Adds missing test coverage to ensure the 6× multiplier for the ultra-premium model claude-opus-4.6-1m is correctly applied all the way through the live-session “Est. Cost” rendering pipeline (preventing regressions where the multiplier is accidentally capped at 3× via tier-based logic).

Changes:

• Add a render_live_sessions integration test asserting claude-opus-4.6-1m displays ~30 for 5 calls and does not display ~15.
• Add a direct unit test for _estimate_premium_cost verifying claude-opus-4.6-1m uses the raw 6.0 multiplier across multiple call counts.