diff --git a/src/libslic3r/CadDocument.cpp b/src/libslic3r/CadDocument.cpp
index 5454f8700f..b6787b9923 100644
--- a/src/libslic3r/CadDocument.cpp
+++ b/src/libslic3r/CadDocument.cpp
@@ -151,6 +151,12 @@ static TopoDS_Wire make_thread_profile(const gp_Pnt& origin, const gp_Dir& xdir,
                                        const gp_Dir& zdir, double radius,
                                        double pitch, double depth, bool internal)
 {
+    // `internal` is intentionally unused: the V is the same shape either way, always pointing
+    // radially outward from `radius`. What differs is what the caller DOES with the swept solid
+    // — an external thread fuses it onto the shaft, an internal one cuts it out of the wall after
+    // boring at the minor diameter. Keeping the parameter documents that the caller decided, and
+    // stops someone "fixing" the profile to point inward for the internal case, which would make
+    // the groove sweep already-empty bore space and cut nothing.
     (void)internal;
     gp_Vec vx(xdir), vz(zdir);
     // Root the V CLEARLY inside the wall (a real overlap, not a 0.05 mm tangency) so the boolean
diff --git a/tests/libslic3r/test_caddocument.cpp b/tests/libslic3r/test_caddocument.cpp
index ca2a4e3204..d2bbf2e144 100644
--- a/tests/libslic3r/test_caddocument.cpp
+++ b/tests/libslic3r/test_caddocument.cpp
@@ -967,11 +967,17 @@ TEST_CASE("extrude taper + up-to-face distance", "[CadDocument]")
     }
 }
 
-// [known-broken]: pre-existing failure, the cut groove volume does not meet the asserted
-// threshold. Excluded from the delegated dev loop so a green run means "I broke nothing",
-// and [NotWorking] excludes it from CI's ctest gate for the same reason as the case above.
-// Tracked in the issue tracker instead — see snaporca-kzy.
-TEST_CASE("internal thread cuts a visible groove into the bore wall", "[CadDocument][known-broken][NotWorking]")
+// Was [known-broken] until the numbers were actually measured (snaporca-kzy). The geometry
+// was right all along; the TEST compared against the wrong reference. An internal thread bores
+// at the MINOR radius (radius - depth) and then carves the groove out to radius + depth, so a
+// tapped hole keeps the crests between turns and therefore holds MORE material than a plain
+// clearance hole at the nominal radius. Asserting the thread removes more than a plain Ø12 bore
+// asked for something no real tapped hole does. Measured on this fixture: plain Ø12 bore removes
+// 2261 mm3, the thread removes 2157 = 1571 (minor bore) + 586 (groove). Exact BRepGProp volume
+// agrees with the tessellated volume to 2.5 mm3, so this was never a meshing artefact either.
+// The tap-drill bore is the honest reference: against it, the groove's 586 mm3 is the thing worth
+// asserting, because that is what "the thread actually cuts" means.
+TEST_CASE("internal thread cuts a visible groove into the bore wall", "[CadDocument][thread]")
 {
     using namespace Slic3r;
     SketchPlane xy = SketchPlane::XY();
@@ -992,10 +998,11 @@ TEST_CASE("internal thread cuts a visible groove into the bore wall", "[CadDocum
         doc.features.push_back(ex);
     };
 
-    // Reference: box with a plain Ø12 bore through it.
+    // Reference: box bored at the MINOR diameter — the tap drill the thread starts from.
+    // Ø10 = 2 * (radius 6 - depth 1), matching what apply_thread cuts before the groove.
     CadDocument hole_doc;
     make_box(hole_doc);
-    hole_doc.add_hole(12.0, 20.0, true, 0.0, 0.0, xy, "Hole");
+    hole_doc.add_hole(10.0, 20.0, true, 0.0, 0.0, xy, "Hole");
     REQUIRE(hole_doc.recompute());
     REQUIRE(hole_doc.error.empty());
     const double v_hole = double(hole_doc.display_mesh.volume());
@@ -1008,10 +1015,9 @@ TEST_CASE("internal thread cuts a visible groove into the bore wall", "[CadDocum
     REQUIRE(thr_doc.error.empty());
     const double v_thread = double(thr_doc.display_mesh.volume());
 
-    // The helical groove must carve material OUT of the wall, beyond the plain
-    // bore -> a visible internal thread. The old inward-pointing profile only
-    // swept already-empty bore space and removed essentially nothing, so it would
-    // give v_thread ~= v_hole; the fixed profile removes a meaningful volume.
+    // The groove must carve material out of the wall BEYOND the tap-drill bore, which is what
+    // makes the thread visible. A profile that only swept already-empty bore space would give
+    // v_thread ~= v_hole; the real one removes several hundred mm3 more.
     REQUIRE(v_thread > 0.0);
     REQUIRE(v_thread < v_hole);
     REQUIRE((v_hole - v_thread) > 20.0);