perf: optimize Java tracing agent (E2E reduction + serialization + writes) #80

Job	Run time
determine-changes	9s
e2e-java (java-fibonacci-nogit, end_to_end_test_java_fibonacci.py, 70, true)	7m 9s
prek	17s
e2e-java (java-void-optimization-nogit, end_to_end_test_java_void_optimization.py, 70, true)	8m 0s
e2e-python (tracer-replay, end_to_end_test_tracer_replay.py, 10)	9m 12s
unit-tests (ubuntu-latest, 3.12)	7m 11s
unit-tests (ubuntu-latest, 3.11)	7m 20s
unit-tests (ubuntu-latest, 3.9)	9m 11s
e2e-python (topological-sort, end_to_end_test_topological_sort_worktree.py, 5)	6m 32s
e2e-python (coverage-e2e, end_to_end_test_coverage.py, black)	7m 53s
unit-tests (ubuntu-latest, 3.13)	8m 19s
e2e-java (java-tracer, end_to_end_test_java_tracer.py, 10)	9m 12s
unit-tests (windows-latest, 3.13)	9m 19s
unit-tests (ubuntu-latest, 3.14)	6m 14s
e2e-python (bubble-sort-unittest, end_to_end_test_bubblesort_unittest.py, 40)	9m 10s
e2e-python (init-optimization, end_to_end_test_init_optimization.py, 10)	7m 47s
e2e-python (bubble-sort-pytest-nogit, end_to_end_test_bubblesort_pytest.py, 70, true)	9m 12s
e2e-python (async-optimization, end_to_end_test_async.py, 10)	9m 12s
e2e-python (futurehouse-structure, end_to_end_test_futurehouse.py, 5)	6m 22s
unit-tests (ubuntu-latest, 3.10)	8m 28s
e2e-python (benchmark-bubble-sort, end_to_end_test_benchmark_sort.py, 5)	9m 11s
e2e-js	-1s
type-check	-1s
required checks passed	2s
	2h 35m 20s

Provide feedback