diff --git a/.github/workflows/c6-factorial-moment.yml b/.github/workflows/c6-factorial-moment.yml
index 1612b7e70..8e037224f 100644
--- a/.github/workflows/c6-factorial-moment.yml
+++ b/.github/workflows/c6-factorial-moment.yml
@@ -15,7 +15,7 @@ permissions:
 jobs:
   exact-and-probe:
     runs-on: ubuntu-latest
-    timeout-minutes: 10
+    timeout-minutes: 20
     steps:
       - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1
         with:
diff --git a/tests/test_c6_factorial_job_budget.py b/tests/test_c6_factorial_job_budget.py
new file mode 100644
index 000000000..c2ab0b091
--- /dev/null
+++ b/tests/test_c6_factorial_job_budget.py
@@ -0,0 +1,94 @@
+"""Keep the factorial job finite without dropping either full regression mode.
+
+Run 37463704292/1 completed its normal 243-test suite in 317.516 seconds,
+then was cancelled during optimized discovery near its 10-minute job limit.
+The 20-minute envelope is an operational budget, not a runtime guarantee or
+permission to increase the unchanged 300-second numerical child limit.
+These tests inspect the current restricted workflow layout, not GitHub's scheduler.
+"""
+from pathlib import Path
+import ast
+import re
+import unittest
+
+ROOT = Path(__file__).resolve().parents[1]
+WORKFLOW = ROOT / '.github/workflows/c6-factorial-moment.yml'
+REGRESSION = 'Full repository regression suite in both Python modes'
+NUMERICAL = 'Verify sources, replay both modes, reject mutants'
+
+
+def named_step(text, name):
+    marker = '      - name: ' + name + '\n'
+    if text.count(marker) != 1:
+        raise ValueError('expected one named step: ' + name)
+    return text.split(marker, 1)[1].split('      - ', 1)[0]
+
+
+def run_block(step):
+    marker = '        run: |\n'
+    if step.count(marker) != 1:
+        raise ValueError('expected one literal run block')
+    lines = step.split(marker, 1)[1].splitlines()
+    if not lines or any(not line.startswith('          ') for line in lines):
+        raise ValueError('unsupported run-block indentation')
+    return '\n'.join(line[10:] for line in lines) + '\n'
+
+
+class FactorialJobBudgetTests(unittest.TestCase):
+    @classmethod
+    def setUpClass(cls):
+        cls.text = WORKFLOW.read_text(encoding='utf-8')
+
+    def test_one_explicit_finite_job_budget(self):
+        # Do not accidentally accept a step timeout, default 360, or expression.
+        limits = re.findall(r'^\s*timeout-minutes:.*$', self.text, re.MULTILINE)
+        self.assertEqual(limits, ['    timeout-minutes: 20'])
+        self.assertIn('  exact-and-probe:\n    runs-on: ubuntu-latest\n'
+                      '    timeout-minutes: 20\n    steps:\n', self.text)
+
+    def test_both_complete_regression_modes_are_unconditional(self):
+        step = named_step(self.text, REGRESSION)
+        self.assertEqual(step,
+                         '        run: |\n'
+                         '          set -euo pipefail\n'
+                         '          python -B -S -m unittest discover -s tests -v\n'
+                         '          python -B -O -S -m unittest discover -s tests -v\n')
+        self.assertLess(self.text.index(REGRESSION), self.text.index(NUMERICAL))
+        for bypass in ('continue-on-error:', 'if:', 'background:'):
+            self.assertNotIn(bypass, self.text)
+
+    def test_numerical_budget_inventory_and_postflight_remain(self):
+        block = run_block(named_step(self.text, NUMERICAL))
+        self.assertTrue(block.startswith("set -euo pipefail\npython -B -S - <<'PY'\n"))
+        self.assertTrue(block.endswith('PY\ngit diff --exit-code\n'))
+        program = block.split("<<'PY'\n", 1)[1].rsplit('\nPY\n', 1)[0]
+        module = ast.parse(program)
+        calls = [n for n in ast.walk(module) if isinstance(n, ast.Call)
+                 and isinstance(n.func, ast.Name) and n.func.id == 'replay']
+        self.assertEqual(len(calls), 1)
+        call = calls[0]
+        self.assertEqual([(kw.arg, ast.literal_eval(kw.value)) for kw in call.keywords],
+                         [('timeout', 300)])
+        loops = [n for n in module.body if isinstance(n, ast.For)
+                 and isinstance(n.target, ast.Name) and n.target.id == 'flags']
+        self.assertEqual(len(loops), 1)
+        self.assertEqual(ast.literal_eval(loops[0].iter), (['-B', '-S'], ['-B', '-O', '-S']))
+        mutants = next(n.value for n in module.body if isinstance(n, ast.Assign)
+                       and any(isinstance(t, ast.Name) and t.id == 'mutants' for t in n.targets))
+        self.assertEqual(ast.literal_eval(mutants),
+                         ['wrong-elimination', 'touching-balls', 'no-log', 'short-remainder'])
+        for guard in ('packet tree differs from manifest:', 'flat regular source required:',
+                      'source identity mismatch:', "'mathematical_acceptance': False"):
+            self.assertIn(guard, program)
+
+    def test_new_test_is_covered_and_execution_permissions_are_unchanged(self):
+        self.assertIn("      - 'tests/**'\n", self.text)
+        self.assertIn('permissions:\n  contents: read\n', self.text)
+        self.assertIn('          persist-credentials: false\n          fetch-depth: 0\n', self.text)
+        self.assertIn("          python-version: '3.11.16'\n", self.text)
+        self.assertIn('actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1', self.text)
+        self.assertIn('actions/setup-python@5fda3b95a4ea91299a34e894583c3862153e4b97', self.text)
+
+
+if __name__ == '__main__':
+    unittest.main()