feat(rum): add rum_metrics resource

[michael-richey]

Summary

First RUM resource in the RUM initiative. Adds rum_metrics (RUM-based metrics, /api/v2/rum/config/metrics), modeled on spans_metrics: skip_resource_mapping=True, full CRUD, and a 409-on-create fallback that hydrates state.destination via GET-by-id then PATCHes — so a first run against a pre-populated destination does not fail permanently on a metric-name conflict.

Stacked on top of #714 (the rum_applications prep PR).

Changes

• datadog_sync/model/rum_metrics.py (new) — RUMMetrics model, resource_type="rum_metrics", base_path="/api/v2/rum/config/metrics", skip_resource_mapping=True.
• datadog_sync/models/__init__.py — register RUMMetrics.
• tests/unit/test_rum_metrics.py (new) — 8 unit tests pinning get_resources, import_resource (by-id + passthrough), create_resource, update_resource, delete_resource, and the 409 fallback contract (GET-by-id → hydrate state → PATCH) plus non-409 reraise.
• README.md — add rum_metrics to the supported-resources and dependency tables (no dependencies).

Testing

• pytest tests/unit/test_rum_metrics.py → 8 passed
• pytest tests/unit/test_rum_metrics.py tests/unit/test_rum_applications.py → 15 passed
• Full unit suite: no new failures (5 pre-existing subprocess-test failures unchanged).

Notes / follow-up

Integration tests + VCR cassettes are deferred — recording requires sandbox-org API access not available in this session. The model behavior is fully covered by unit tests (AsyncMock clients, same pattern as test_spans_metrics_409_fallback.py). A follow-up will add tests/integration/resources/test_rum_metrics.py (subclassing BaseResourcesTestClass, field_to_update="attributes.filter.query") and record the 8 cassettes.

[michael-richey]

Solid addition.

Highlights:

• Model implementation is consistent with existing spans_metrics semantics.
• The POST-409 fallback to GET-by-id + PATCH is a strong reliability pattern for skip_resource_mapping=True resources.
• Tests cover happy path plus non-409 re-raise behavior, which is exactly what we want.

No blocking issues found in this PR diff.

[Copilot]

Copilot review overview

🟢 Approval recommended

The implementation follows established resource patterns and has focused unit coverage for its key behaviors.

Review effort: Balanced
Findings: None

What changed in this PR

Adds rum_metrics synchronization with full CRUD support and conflict recovery for pre-existing destination metrics.

Changes:

• Implements and registers the RUMMetrics resource model.
• Adds unit coverage for CRUD and 409 fallback behavior.
• Documents rum_metrics as a supported, dependency-free resource.

File	Description
datadog_sync/​model/​rum_metrics.py	Implements RUM metric synchronization.
datadog_sync/​models/​__init__.py	Registers the new resource model.
tests/​unit/​test_rum_metrics.py	Covers CRUD and conflict handling.
README.md	Documents resource support and dependencies.

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

[michael-richey]

Nice addition overall. The model/test shape is consistent with spans_metrics, and the 409 fallback contract is clearly tested.

A couple things I liked in particular:

• The POST(409) -> GET-by-id -> PATCH fallback is explicit and covered.
• Help-text/documentation updates around import directionality are clear and operator-friendly.

No blocking issues from my side on this PR.