fix: harden hostname tracking release behavior

CONTEXT.md

@@ -56,7 +56,7 @@ When a change alters one of these boundaries, update this document and the neare
 
 ### Deferred boundaries
 
-- Cross-run change detection, alerts, and automatic reactions remain separate future product decisions. A scheduled occurrence only submits finite enumeration runs.
+- Read-only cross-run change projections over finalized evidence are part of the release contract. Alerts and automatic reactions remain deferred, and a scheduled occurrence only submits finite enumeration runs.
 - Cross-run source ranking remains a reporting decision. One run's hostname-yield summary does not automatically select, disable, or rank sources.
 - Distributed workers, multi-host operation, PostgreSQL, and hosted multi-user authorization require measured demand and new decisions. The release remains SQLite-first and local-operator focused.
 - Automatic scope expansion remains deferred. Evidence can suggest a later target, while the operator controls every scope change.

tests/e2e/test_harvestview.py

@@ -533,6 +533,14 @@ def test_hostname_tracking_filters_persisted_run_changes(
     expect(panel).to_contain_text('TimeoutError')
     expect(panel).not_to_contain_text('stable.example.test')
 
+    page.locator('#tracking-change-filter').select_option('persisting')
+    expect(page.locator('#tracking-persisting-filter')).to_be_checked()
+    expect(rows).to_have_count(1)
+    expect(panel).to_contain_text('stable.example.test')
+    page.locator('#tracking-persisting-filter').uncheck()
+    expect(page.locator('#tracking-change-filter')).to_have_value('')
+    expect(rows).to_have_count(3)
+
     page.locator('#tracking-exclusive-filter').check()
     expect(rows).to_have_count(2)
     expect(panel).not_to_contain_text('missing.example.test')

tests/test_source_yields_cli.py

@@ -16,6 +16,7 @@ from theHarvester.lib.database import ResultStore
 from theHarvester.lib.evidence_types import RESULT_KINDS, ExecutionStatus
 
 RUN_ONE = UUID('11111111-1111-4111-8111-111111111111')
+RUN_TIE_LATER = UUID('11111111-1111-4111-8111-111111111112')
 RUN_TWO = UUID('22222222-2222-4222-8222-222222222222')
 
 
@@ -572,6 +573,42 @@ def test_target_changes_reports_every_run_pair_and_a_clear_null_baseline(
     assert {row['run_id'] for row in payload['hostname_changes']} == {str(RUN_TWO)}
 
 
+def test_run_id_breaks_an_equal_completion_time_tie_deterministically(
+    tmp_path: Path,
+    capsys: pytest.CaptureFixture[str],
+) -> None:
+    database = tmp_path / 'runs.sqlite'
+    store = ResultStore(database)
+    asyncio.run(store.initialize())
+    asyncio.run(
+        store.save_run(
+            _completed_run(
+                RUN_TIE_LATER,
+                observations=(ResultObservation('alpha', 'hostname', 'later.example.test'),),
+            )
+        )
+    )
+    asyncio.run(
+        store.save_run(
+            _completed_run(
+                RUN_ONE,
+                observations=(ResultObservation('alpha', 'hostname', 'earlier.example.test'),),
+            )
+        )
+    )
+    asyncio.run(store.dispose())
+
+    assert (
+        source_yields.main(
+            ['--database', str(database), '--run-id', str(RUN_TIE_LATER), '--changes', '--format', 'json']
+        )
+        == 0
+    )
+
+    comparison = json.loads(capsys.readouterr().out)['comparisons'][0]
+    assert comparison['baseline_run_id'] == str(RUN_ONE)
+
+
 def test_include_persisting_adds_unchanged_hostname_rows_without_changing_counts(
     tmp_path: Path,
     capsys: pytest.CaptureFixture[str],
@@ -729,6 +766,9 @@ def test_unscoped_empty_database_reports_an_explicit_empty_scope(
     assert source_yields.main(['--database', str(database)]) == 0
     assert capsys.readouterr().out.splitlines()[:3] == ['Targets: none', 'Kind: hostname', 'Run count: 0']
 
+    assert source_yields.main(['--database', str(database), '--changes']) == 0
+    assert capsys.readouterr().out.startswith('Targets: none\nComparison count: 0\n')
+
 
 def test_missing_database_fails_without_creating_file(
     tmp_path: Path,

theHarvester/lib/api/static/harvestview/app.js

@@ -1422,10 +1422,19 @@
     const query = event.target.value.trim().toLowerCase();
     state.resultTable?.setFilter(row => !query || row.value.toLowerCase().includes(query));
   });
-  for (const filter of [
-    nodes.trackingChange, nodes.trackingSource, nodes.trackingResolution,
-    nodes.trackingExclusive, nodes.trackingPersisting
-  ]) filter.addEventListener('change', () => renderHostnameTracking(state.detail));
+  nodes.trackingChange.addEventListener('change', () => {
+    if (nodes.trackingChange.value === 'persisting') nodes.trackingPersisting.checked = true;
+    renderHostnameTracking(state.detail);
+  });
+  nodes.trackingPersisting.addEventListener('change', () => {
+    if (!nodes.trackingPersisting.checked && nodes.trackingChange.value === 'persisting') {
+      nodes.trackingChange.value = '';
+    }
+    renderHostnameTracking(state.detail);
+  });
+  for (const filter of [nodes.trackingSource, nodes.trackingResolution, nodes.trackingExclusive]) {
+    filter.addEventListener('change', () => renderHostnameTracking(state.detail));
+  }
   nodes.sourceSearch.addEventListener('input', event => renderSourceGroups(event.target.value));
   nodes.selectCapability.addEventListener('click', selectCapability);
   nodes.selectP0.addEventListener('click', () => setP0Selection(true));

theHarvester/source_yields.py

@@ -210,7 +210,8 @@ def _targets_table(rows: list[dict[str, str | int]]) -> str:
 def _tracking_table(payload: dict[str, object]) -> str:
     comparisons = cast('list[dict[str, object]]', payload['comparisons'])
     changes = cast('list[dict[str, object]]', payload['hostname_changes'])
-    lines = [f'Target: {payload["target"]}', f'Comparison count: {payload["comparison_count"]}']
+    scope = f'Target: {payload["target"]}' if payload['target'] is not None else 'Targets: none'
+    lines = [scope, f'Comparison count: {payload["comparison_count"]}']
     summary_columns = (
         ('run_id', 'CURRENT RUN'),
         ('baseline_run_id', 'BASELINE RUN'),
@@ -227,7 +228,7 @@ def _tracking_table(payload: dict[str, object]) -> str:
         }
         for comparison in comparisons
     ]
-    summary_widths = [max(len(label), *(len(str(row[key])) for row in summary_rows)) for key, label in summary_columns]
+    summary_widths = [max([len(label), *(len(str(row[key])) for row in summary_rows)]) for key, label in summary_columns]
     lines.append('  '.join(label.ljust(summary_widths[index]) for index, (_key, label) in enumerate(summary_columns)))
     lines.extend(
         '  '.join(str(row[key]).ljust(summary_widths[index]) for index, (key, _label) in enumerate(summary_columns))