fix: add depth guard to sync_node_execution to prevent unbounded recursion

[falloficaruss]

Tracking issue

Closes flyteorg/flyte#7338

Why are the changes needed?

FlyteRemote.sync_node_execution recursively syncs launched-LP executions and parent-node children with no depth limit. A workflow graph with
deeply nested launched LPs triggers RecursionError, crashing the Python process. Every pyflyte run --remote call against such an execution
blows up.

What changes were proposed in this pull request?

• Added _depth and _max_depth parameters to _sync_execution and sync_node_execution (default _max_depth=50)
  • Both methods raise FlyteAssertion at entry when _depth > _max_depth instead of overflowing the Python call stack
  • Updated all 6 recursive call sites to pass _depth + 1: the per-node sync in _sync_execution, the launched-LP path, dynamic subworkflow
    children, static subworkflow children, and both branch-node child paths

How was this patch tested?

• Added tests/flytekit/unit/remote/test_recursion_guard.py with 4 tests:
  • test_max_depth_raises_flyte_assertion — verifies FlyteAssertion is raised when depth exceeds limit
  • test_reasonable_depth_does_not_raise — verifies default depth allows normal operation
  • test_nested_under_default_limit — verifies nested calls under the limit succeed
  • test_sync_execution_depth_guard — verifies the guard on _sync_execution independently
  • All 44 existing tests in tests/flytekit/unit/remote/ pass with no regressions

Setup process

Screenshots

Check all the applicable boxes

• I updated the documentation accordingly.
• All new and existing tests passed.
• All commits are signed-off.

Related PRs

Docs link