Surface XGEN equivalence diagnostics

[SonAIengine]

Summary

• include target_equivalence_groups and target_equivalence_group_count in XGEN deterministic target selector diagnostics
• include equivalence group counts in built-in XGEN fixture summaries and top-level suite summary
• include equivalence group fields in XGEN API scale benchmark case/search artifacts
• document the built-in suite diagnostic artifact and counts

Why

PR #46 exposes product-neutral tool equivalence evidence. This PR wires that evidence into the XGEN-style benchmark artifacts so adapter/UI/SSE work can consume it from the same target selector diagnostics path instead of recomputing it.

Candidate selection behavior is unchanged. The new fields are diagnostic/evidence only.

Evidence

Built-in deterministic suite:

poetry run python -m benchmarks.xgen_tool_graph.run --suite all --json > /tmp/gtc-xgen-equivalence-diagnostics.json

Observed top-level summary:

• avg_target_equivalence_group_count: 0.333333
• target_equivalence_group_case_count: 5

Validation

• poetry run pytest tests/test_xgen_tool_graph_benchmark.py tests/test_xgen_api_scale_benchmark.py -q
• poetry run ruff check .
• poetry run ruff format --check .
• poetry run pytest tests/ -q (957 passed, 3 skipped)

[SonAIengine]

Closing this internal stacked PR as superseded by the current main branch. The branch is preserved for history.