test(engine): cover Bloodthirsty Tenacity loop

[mike-theDude]

Summary

• Add a live-loop regression test for Bloodthirsty Conqueror + Enduring Tenacity.
• Verify Enduring Tenacity's targeted trigger auto-resolves to the sole opponent.
• Pin Bloodthirsty Conquerer x Enduring Tenacity Inifinite Loop — This loop isn't triggering for the easy runaway when combo… #4787 as already fixed on current main by proving the combo wins through the persisted loop ring.

Closes #4787

Validation

• cargo fmt --all -- --check
• cargo test -p engine drive_bloodthirsty_conqueror_enduring_tenacity_wins_live -- --nocapture
• cargo clippy -p engine --tests -- -D warnings

QA

Final recommendation: SCOPE READY

Tilt was unavailable in the isolated worktree:
No tilt apiserver found: tilt-default

Used direct cargo fallback per repo policy. Full engine suite was not run.

Gate A

Gate A PASS head=ee3abc893aa72873f7d5db4f91470d15c74759e1 base=7e1ddc9b778c3c3bcfa9e56d1f726561225c099d

Anchored on

• crates/engine/src/analysis/corpus_tests.rs:966 — existing targeted drain/refill loop witness for Sanguine Bond + Exquisite Blood.
• crates/engine/src/analysis/corpus_tests.rs:974 — existing live-loop test structure using build_drain_board, seed_lifegain_cascade, drive_pass_priority, and first_gameover_beat.

CR annotations

No runtime rules logic changed; test-only regression guard.

Model: gpt-5
Thinking: high
Tier: Standard

[gemini-code-assist]

Warning

You have reached your daily quota limit. Please wait up to 24 hours and I will start processing your requests again!

[github-actions]

Parse changes introduced by this PR

✓ No card-parse changes detected.

[matthewevans]

APPROVED

🔴 Blocker

• None.

🟡 Non-blocking

• None.

✅ Clean

• The exact reported card pair runs through the real pass-priority corpus pipeline, verifies the targeted trigger auto-resolves for the sole opponent, and requires an early game-over at 200 life; reverting the loop shortcut would fail this regression.
• Current-head CI, including Rust lint/tests and card-data validation, is green; the parse-diff artifact reports no card-parse changes.

Recommendation: approve and enqueue as a high-value test regression.