chore(governance): add dispatch benchmark task for nine-skill coverage

- Add B024 golden task for reflective-dispatch routing
- Sync QUALITY_GATES_SUMMARY counts and nine-skill wording
- Anti-drift: benchmark must cover all VALID_WORKFLOWS

Author: johnteee

reflective-prompt-library/plans/QUALITY_GATES_SUMMARY.md

@@ -239,16 +239,16 @@ python3 reflective-prompt-library/plans/validate_route_fixture.py
 **File:** `reflective-prompt-library/plans/benchmark_tasks.py`
 
 **What it does:**
-- Defines 23 golden tasks for validation
-- Covers all major benchmarked workflows (8 different skills)
-- Balanced difficulty distribution (5 easy, 11 medium, 7 hard)
+- Defines 24 golden tasks for validation
+- Covers all nine frozen workflow skills
+- Balanced difficulty distribution (6 easy, 11 medium, 7 hard)
 - Diverse categories (15 different categories)
 - Clear acceptance criteria for each task
 - Expected workflow mapping
 
 **Results:**
-- 23 benchmark tasks created
-- Tasks cover: implementation, planning, review, research, risk, handoff, debugging, refactoring, retrospective, runtime governance, scaffold provenance, SOP compiler planning, minimality, test planning, and workflow design
+- 24 benchmark tasks created
+- Tasks cover: routing, implementation, planning, review, research, risk, handoff, debugging, refactoring, retrospective, runtime governance, scaffold provenance, SOP compiler planning, minimality, test planning, and workflow design
 - Output: `plans/benchmark-tasks.json`
 
 **Usage:**
@@ -277,7 +277,7 @@ The implementation aligns with research findings:
 | Files validated | 103 | ✅ All pass |
 | Links validated | 0 broken | ✅ Perfect |
 | Skills with governance | 9/9 | ✅ Complete |
-| Benchmark tasks | 23 | ✅ Ready |
+| Benchmark tasks | 24 | ✅ Ready |
 | Routing consistency | 100.0% | ✅ Passes ROUTE-001 expanded boundary eval |
 | Holdout routing consistency | 100.0% | ✅ ROUTE-002 (36 groups, 100 paraphrases) |
 | Adversarial routing consistency | 100.0% | ✅ ROUTE-003 (13 groups, 31 paraphrases) |

reflective-prompt-library/plans/benchmark_tasks.py

@@ -379,6 +379,21 @@ def define_tasks(self) -> List[BenchmarkTask]:
                     "Side effects define authority, idempotency, retry caps, compensation, and hard stops",
                     "Observability and scenario tests separate specified behavior from unproven runtime guarantees"
                 ]
+            ),
+            BenchmarkTask(
+                id="B024",
+                name="Workflow skill selection for mixed intent",
+                description="Classify a mixed user request and select the smallest useful reflective workflow skill with explicit route trace",
+                expected_workflow="reflective-dispatch",
+                difficulty="easy",
+                category="routing",
+                acceptance_criteria=[
+                    "Primary workflow skill is named with rationale tied to intent signals",
+                    "Strictness level is stated with risk and cost calibration",
+                    "Route trace includes confidence and at least one rejected alternate route",
+                    "Smallest sufficient workflow is chosen over plan-orchestration bloat",
+                    "Next action is explicit and matches the selected workflow"
+                ]
             )
         ]
 

reflective-prompt-library/plans/tests/test_quality_gates_summary.py

@@ -104,3 +104,9 @@ def test_research_alignment_lists_nine_frozen_skills(summary_text: str):
     assert "nine frozen workflow skills" in section
     assert "8 lifecycle skills" not in section
 
+def test_benchmark_section_covers_nine_workflows(summary_text: str):
+    section = summary_text.split("### 7. Small Benchmark Set", 1)[1].split("## Research Alignment", 1)[0]
+    assert "nine frozen workflow skills" in section
+    assert "8 different skills" not in section
+    assert "24 golden tasks" in section or "24 benchmark tasks" in section
+

reflective-prompt-library/plans/tests/test_validate_benchmark_fixture.py

@@ -23,3 +23,11 @@ def test_every_task_maps_to_known_workflow():
 def test_benchmark_ids_are_unique():
     ids = [task.id for task in BenchmarkSet().tasks]
     assert len(ids) == len(set(ids))
+
+def test_benchmark_covers_all_nine_workflows():
+    workflows = {task.expected_workflow for task in BenchmarkSet().tasks}
+    assert workflows == VALID_WORKFLOWS, (
+        f"missing workflows: {sorted(VALID_WORKFLOWS - workflows)}; "
+        f"unexpected: {sorted(workflows - VALID_WORKFLOWS)}"
+    )
+