eval: preflight execution probe label coverage

[AbdelStark]

Parent

Parent #385
Closes #390

Summary

• add a schema-versioned execution-probe label coverage preflight before latent extraction
• emit typed blockers and a not-evaluable latent-probe report when requested targets lack train/val/test coverage
• add a multi-seed representation gate table helper for pass/fail and magnitude target summaries
• fix latent probes so boolean False labels count as real labels instead of missing values
• document coverage reports and report-template handling for v0.9 representation gates

Validation

• uv run pytest tests/eval/test_latent_probe.py tests/eval/test_execution_probe_coverage.py tests/eval/test_execution_probe_targets.py tests/eval/test_execution_eval_cli.py
• uv run python -m compileall -q codelewm/eval/latent_probe.py codelewm/eval/execution_probe_coverage.py codelewm/eval/execution_runner.py tests/eval/test_latent_probe.py tests/eval/test_execution_probe_coverage.py
• uv run python -m compileall -q -x 'tests/fixtures/codestate/invalid_(before|after)\.py$' codelewm tests scripts/build-passfail-pack
• git diff --check
• uv run pytest tests/ (955 passed, 8 skipped, 1 warning)