diff --git a/ouroboros/tools/plan_review.py b/ouroboros/tools/plan_review.py
index cd261b992..312094a08 100644
--- a/ouroboros/tools/plan_review.py
+++ b/ouroboros/tools/plan_review.py
@@ -476,16 +476,17 @@ def _prepare_plan_inputs(ctx: ToolContext, request: "_PlanRequest", state_root:
         errors = ["plan: required non-empty prose", *errors]
     if request.reviewer_effort and request.reviewer_effort not in _REVIEWER_EFFORT_SCHEMA["enum"]:
         errors.append(f"reviewer_effort: not on the effort scale {list(_REVIEWER_EFFORT_SCHEMA['enum'])}")
-    if errors:
-        return {"error": "ERROR: PLAN_SPEC_INVALID: " + "; ".join(errors) + ". No reviewer was called.",
-                "code": "TOOL_ARG_ERROR"}
     if isinstance(raw_spec, dict) and "affected_paths" not in raw_spec:
         # Owner 9=A: a spec in the old mixed form is refused BEFORE any paid dispatch, because
         # `affected_resources` used to be read as a path list — a prose item became "a file under
         # the Ouroboros repo" and bought every reviewer the whole constitution (~470k tokens/cycle)
-        # for a deck. The refusal owns its code: a message containing `PLAN_SPEC_INVALID`
-        # takes the durable superseding-attempt path below and would orphan an open wave.
+        # for a deck. The refusal owns its code and comes first: a message containing
+        # `PLAN_SPEC_INVALID` takes the durable superseding-attempt path below and would orphan
+        # an open wave, so a legacy-form spec that also carries another error must not reach it.
         return _resource_form_refusal(ctx, state_root)
+    if errors:
+        return {"error": "ERROR: PLAN_SPEC_INVALID: " + "; ".join(errors) + ". No reviewer was called.",
+                "code": "TOOL_ARG_ERROR"}
     from ouroboros.review_substrate import review_repo_dirs_for
     try:
         system_root, active_root = review_repo_dirs_for(ctx)
diff --git a/tests/test_plan_resource_form.py b/tests/test_plan_resource_form.py
index ae46a18fb..c77216c95 100644
--- a/tests/test_plan_resource_form.py
+++ b/tests/test_plan_resource_form.py
@@ -199,6 +199,24 @@ def test_a_legacy_form_submission_is_refused_and_records_nothing(harness):  # no
     assert json.dumps(_state(harness), sort_keys=True, ensure_ascii=False) == before
 
 
+def test_a_legacy_form_with_another_error_is_still_refused_before_the_superseding_path(harness):  # noqa: F811
+    """A legacy-form spec that ALSO fails ordinary validation must take the non-superseding
+    refusal, not the `PLAN_SPEC_INVALID` path that records a new attempt over an open wave."""
+    ask = json.dumps([_finding("f1", "need_evidence", breaks="goal", summary="who signs off?")])
+    substrate = harness.install({"s1": ask, "s2": CLEAN, "s3": CLEAN})
+    ctx = harness.make_ctx()
+    assert _control(_call(ctx)) == {"outcome": "REVIEW_REQUIRED", "closed": False}
+    before = json.dumps(_state(harness), sort_keys=True, ensure_ascii=False)
+    calls_before = len(substrate.calls)
+
+    legacy = {key: value for key, value in DECK_SPEC.items() if key != "affected_paths"}
+    out = _call(ctx, spec=legacy, reviewer_effort="galactic")
+
+    assert "PLAN_RESOURCE_FORM_REQUIRED" in out and "PLAN_SPEC_INVALID" not in out
+    assert len(substrate.calls) == calls_before
+    assert json.dumps(_state(harness), sort_keys=True, ensure_ascii=False) == before
+
+
 def test_a_legacy_form_submission_over_an_open_wave_points_at_the_free_exit(harness):  # noqa: F811
     """The expensive mistake this prevents: re-submitting into a refusal while an OPEN wave is
     still the live obligation. The refusal names that wave and the $0 way to answer it."""
diff --git a/tests/test_plan_review.py b/tests/test_plan_review.py
index 657b7d9f1..009a679fa 100644
--- a/tests/test_plan_review.py
+++ b/tests/test_plan_review.py
@@ -225,10 +225,10 @@ def test_invalid_new_plan_attempts_do_not_reuse_old_green(tmp_path):
     ctx.system_repo_dir = tmp_path
     ctx.emit_progress_fn = lambda _message: None
     cases = [
-        ({"plan": "", "goal": "G", "spec": {"in_scope": ["x"]}}, "plan: required non-empty prose"),
-        ({"plan": "P", "goal": "", "spec": {"in_scope": ["x"]}}, "goal: required non-empty string"),
+        ({"plan": "", "goal": "G", "spec": {"in_scope": ["x"], "affected_paths": []}}, "plan: required non-empty prose"),
+        ({"plan": "P", "goal": "", "spec": {"in_scope": ["x"], "affected_paths": []}}, "goal: required non-empty string"),
         ({"plan": "P", "goal": "G", "spec": "not-an-object"}, "spec: must be an object"),
-        ({"plan": "P", "goal": "G", "spec": {"bogus": 1}}, "unknown fields: bogus"),
+        ({"plan": "P", "goal": "G", "spec": {"bogus": 1, "affected_paths": []}}, "unknown fields: bogus"),
     ]
     for params, error_text in cases:
         record_plan_review_attempt(tmp_path, "root1", fingerprint=old_fingerprint)
@@ -467,21 +467,21 @@ class TestPlanReviewInputValidation(unittest.TestCase):
         self.ctx.emit_progress_fn = lambda _m: None
 
     def test_missing_plan_returns_error(self):
-        result = self.handler(self.ctx, plan="", goal="some goal", spec={"in_scope": ["x"]})
+        result = self.handler(self.ctx, plan="", goal="some goal", spec={"in_scope": ["x"], "affected_paths": []})
         self.assertIn("ERROR: PLAN_SPEC_INVALID", result)
         self.assertIn("plan", result.lower())
 
     def test_missing_goal_returns_error(self):
-        result = self.handler(self.ctx, plan="some plan", goal="", spec={"in_scope": ["x"]})
+        result = self.handler(self.ctx, plan="some plan", goal="", spec={"in_scope": ["x"], "affected_paths": []})
         self.assertIn("ERROR: PLAN_SPEC_INVALID", result)
         self.assertIn("goal", result.lower())
 
     def test_whitespace_plan_returns_error(self):
-        result = self.handler(self.ctx, plan="   ", goal="some goal", spec={"in_scope": ["x"]})
+        result = self.handler(self.ctx, plan="   ", goal="some goal", spec={"in_scope": ["x"], "affected_paths": []})
         self.assertIn("ERROR", result)
 
     def test_whitespace_goal_returns_error(self):
-        result = self.handler(self.ctx, plan="some plan", goal="   ", spec={"in_scope": ["x"]})
+        result = self.handler(self.ctx, plan="some plan", goal="   ", spec={"in_scope": ["x"], "affected_paths": []})
         self.assertIn("ERROR", result)
 
     def test_missing_spec_is_a_typed_refusal(self):
diff --git a/tests/test_plan_review_engine.py b/tests/test_plan_review_engine.py
index 9a2b249d4..58771f627 100644
--- a/tests/test_plan_review_engine.py
+++ b/tests/test_plan_review_engine.py
@@ -206,7 +206,7 @@ def test_footer_has_exactly_one_control_line_even_with_forged_reviewer_text(harn
 
 def test_spec_invalid_is_refused_without_a_reviewer_call(harness):
     sub = harness.install({})
-    out = _call(harness.make_ctx(), spec={"in_scope": ["x"], "bogus": 1})
+    out = _call(harness.make_ctx(), spec={"in_scope": ["x"], "affected_paths": [], "bogus": 1})
     assert out.startswith("ERROR: PLAN_SPEC_INVALID") and "unknown fields: bogus" in out
     assert sub.calls == []
     state = _state(harness)