feat(measurement): add W&B benchmark identity config

[lipikaramaswamy]

Summary

• Add optional benchmark identity metadata to W&B strict import config and run config.
• Add import_wandb_run.py flags for benchmark role/kind, suite version, branch, PR/release metadata, Anonymizer commit, config ID, and mode.
• Validate that kind=pr and kind=branch include a PR number so candidate benchmark runs can be filtered consistently in W&B.
• Document the benchmark identity fields used by the experiment runner.

Tests

• python3 -m py_compile tools/measurement/import_wandb_run.py tools/measurement/measurement_tools/wandb_models.py
• uv run pytest tests/tools/test_measurement_strict_import_publisher.py tests/tools/test_measurement_wandb_logging.py
• uv run ruff check tools/measurement/import_wandb_run.py tools/measurement/measurement_tools/wandb_models.py tests/tools/test_measurement_strict_import_publisher.py tests/tools/test_measurement_wandb_logging.py
• git diff --check

Notes

The companion anonymizer-experiments branch wires these fields through suite requests and W&B publication.

[greptile-apps]

Greptile Summary

Adds validated benchmark identity metadata to strict W&B imports.

• Adds CLI flags and typed W&B config fields for benchmark role, kind, suite, branch, commit, PR/release, configuration, and mode.
• Enforces compatible role/kind pairs and consistent full/short commit identifiers.
• Refreshes benchmark identity fields when an already-complete import is resumed.
• Preserves launcher-provided WANDB_API_KEY during W&B SDK execution and updates documentation and tests accordingly.

Confidence Score: 5/5

The PR appears safe to merge.

No blocking failure remains.

Important Files Changed

Filename	Overview
tools/measurement/import_wandb_run.py	Builds validated optional benchmark identity metadata from CLI arguments and includes it in sealed-import payloads.
tools/measurement/measurement_tools/wandb_models.py	Defines and validates benchmark identity metadata and projects it into W&B configuration.
tools/measurement/measurement_tools/wandb_setup.py	Refreshes identity fields on completed resumed imports and allows an explicitly inherited W&B API key.
tests/tools/test_measurement_wandb_logging.py	Covers commit consistency, role/kind compatibility, PR requirements, and API-key preservation.
tests/tools/test_measurement_strict_import_publisher.py	Covers identity projection and identity-only refreshes for completed imports.
tools/measurement/README.md	Documents launcher authentication and the new benchmark identity flags.

_{Reviews (7): Last reviewed commit: "fix(measurement): refresh W&B identity o..." | Re-trigger Greptile}