[test] exercise mock workflow from fork

[kaozyn]

Summary

• clarify the README note around CPU-safe local validation before maintainer GPU time
• use this PR to exercise the forked-contributor mock workflow end to end

Validation

• uv run --extra test python -m pytest tests/ strategy/tests/ eval/tests/ -q

Scorecard

RESULT_JSON: {"mode":"mock-workflow","device":"cpu-safe-local","note":"test PR for queue and mock GPU batch"}

<----ia--------e------- relabel refresh: 2026-07-07T14:23:28Z -->

[github-actions]

Gate chain passed. This PR is queued for the next batched GPU evaluation window.

[github-actions]

GPU evaluation complete on RTX 5090 (mock).

• Verdict: eval:BASELINE
  Reason: first admitted strategy on this track
• Track: full-rank / transform mock_transform
• Accuracy: 0.94
• Latency: 0.023 seconds
• Peak VRAM: 1536.0 MiB
• FLOP ratio vs exact: 2.5
• Score: 10.4
• Seed: 3998309171

[github-actions]

Gate chain passed. This PR is queued for the next batched GPU evaluation window.