Expose SmartSeed selected seed outputs (#17)

Adds selected_seed outputs for CADE Easy and MG_SuperSimple while preserving existing output slot order.
Adds regression coverage for output contracts.

README.md

@@ -56,7 +56,7 @@ Photo Dog
 - MGHybrid scheduler: hybrid Karras/Beta sigma stack with smooth tail blending and tiny schedule jitter (ZeSmart-inspired) for more stable, detail-friendly denoising; used by CADE and SuperSimple by default
 - Seed Latent (MG_SeedLatent): fast, deterministic latent initializer aligned to VAE stride; supports pure-noise starts or image-mixed starts (encode + noise) to gently bias content; batch-ready and resolution-agnostic, pairs well with SuperSimple recommended latent sizes for reproducible pipelines
 - Muse Blend and Polish: directional post-mix and final low-frequency-preserving clean-up
-- SmartSeed (CADE Easy and SuperSimple): set `seed = 0` to auto-pick a good seed from a tiny low-step probe. Uses a low-discrepancy sweep, avoids speckles/overexposure, and, if available, leverages CLIP-Vision (with `reference_image`) and CLIPSeg focus text to favor semantically aligned candidates. Logs `Smart_seed_random: Start/End`. 
+- SmartSeed (CADE Easy and SuperSimple): set `seed = 0` to auto-pick a good seed from a tiny low-step probe. Uses a low-discrepancy sweep, avoids speckles/overexposure, and, if available, leverages CLIP-Vision (with `reference_image`) and CLIPSeg focus text to favor semantically aligned candidates. Logs `Smart_seed_random: Start/End` and exposes the final choice on the `selected_seed` output.
 <b>I highly recommend working with SmartSeed.</b>
 - CADE2.5 pipeline does not just upscale the image, it iterates and adds small details, doing it carefully, at every stage.
 
@@ -141,7 +141,7 @@ Notes:
 
 12) Very often, the image in `step 3 is of very good quality`, but it usually lacks sharpness. But if you have a `weak system`, you can `limit yourself to 3 steps`.
 
-13) SmartSeed (auto seed pick): set `seed = 0` in Easy or SuperSimple. The node will sample several candidate seeds and do a quick low‑step probe to choose a balanced one. You’ll see logs `Smart_seed_random: Start` and `Smart_seed_random: End. Seed is: <number>`. Use any non‑zero seed for fully deterministic runs.
+13) SmartSeed (auto seed pick): set `seed = 0` in Easy or SuperSimple. The node will sample several candidate seeds and do a quick low‑step probe to choose a balanced one. You’ll see logs `Smart_seed_random: Start` and `Smart_seed_random: End. Seed is: <number>`, and the chosen value is available from the `selected_seed` output. Use any non‑zero seed for fully deterministic runs.
 
 14) The 4th step sometimes saves the image for a long time, just wait for the end of the process, it depends on the initial resolution you set.
 
@@ -222,6 +222,7 @@ MagicNodes/
 
 ## CADE 2.5 (mg_cade25.py, mg_cade25_easy.py)
 - Deterministic preflight: CLIPSeg pinned to CPU; preview mask reset; noise tied to `iter_seed`
+- CADE Easy appends a `selected_seed` INT output after `mask_preview`; existing LATENT, IMAGE, and mask slots keep their order.
 - Encode/Decode: stride-aligned, with larger overlap for >2K to avoid artifacts
 - Polish mode (final hi-res refinement):
   - `polish_enable`, `polish_keep_low` (global form from reference), `polish_edge_lock`, `polish_sigma`

docs/EasyNodes.md

@@ -40,13 +40,14 @@ Behavior
 - ControlFusion inside this node always relies on presets (no additional CF UI here) to keep the surface minimal.
 
 Outputs
-- `(LATENT, IMAGE)` from the final executed step (e.g., step 2 if `step_count=2`). No preview outputs.
+- `(LATENT, IMAGE, selected_seed)` from the final executed step (e.g., step 2 if `step_count=2`). No preview outputs.
+- `selected_seed` is the seed CADE actually used after SmartSeed selection. With a non-zero fixed seed, it matches the requested seed.
 
 Quickstart
 1) Drop `MG_SuperSimple` into your graph under `MagicNodes/Easy`.
 2) Connect `model/positive/negative/vae/latent`, and a `control_net` module; optionally connect `reference_image` and `clip_vision`.
 3) Choose `step_count` (2/3/4). Leave `custom` Off to use pure presets, or enable it to apply your `seed/steps/cfg/denoise/sampler/scheduler/clipseg_text` across all steps (with Step 1 `denoise=1.0`).
-4) Run. The node returns the final `(LATENT, IMAGE)` for the chosen depth.
+4) Run. The node returns the final `(LATENT, IMAGE, selected_seed)` for the chosen depth.
 
 Notes
 - Presets are read from `pressets/mg_cade25.cfg` and `pressets/mg_controlfusion.cfg`. Keep them in UTF‑8 and prefer `$(ROOT)` over absolute paths.

mod/easy/mg_cade25_easy.py

@@ -2260,8 +2260,8 @@ class ComfyAdaptiveDetailEnhancer25:
             },
         }
 
-    RETURN_TYPES = ("LATENT", "IMAGE", "INT", "FLOAT", "FLOAT", "IMAGE")
-    RETURN_NAMES = ("LATENT", "IMAGE", "steps", "cfg", "denoise", "mask_preview")
+    RETURN_TYPES = ("LATENT", "IMAGE", "INT", "FLOAT", "FLOAT", "IMAGE", "INT")
+    RETURN_NAMES = ("LATENT", "IMAGE", "steps", "cfg", "denoise", "mask_preview", "selected_seed")
     FUNCTION = "apply_cade2"
     CATEGORY = "MagicNodes"
 
@@ -3428,7 +3428,7 @@ class ComfyAdaptiveDetailEnhancer25:
         except Exception:
             pass
 
-        return current_latent, image, int(current_steps), float(current_cfg), float(current_denoise), onnx_mask_img
+        return current_latent, image, int(current_steps), float(current_cfg), float(current_denoise), onnx_mask_img, int(seed)
 
 
 # === Easy UI wrapper: show only top-level controls ===
@@ -3461,8 +3461,8 @@ class CADEEasyUI(ComfyAdaptiveDetailEnhancer25):
         }
 
     # Easy outputs (hide steps/cfg/denoise)
-    RETURN_TYPES = ("LATENT", "IMAGE", "IMAGE")
-    RETURN_NAMES = ("LATENT", "IMAGE", "mask_preview")
+    RETURN_TYPES = ("LATENT", "IMAGE", "IMAGE", "INT")
+    RETURN_NAMES = ("LATENT", "IMAGE", "mask_preview", "selected_seed")
     FUNCTION = "apply_easy"
 
     def apply_easy(self,
@@ -3470,7 +3470,7 @@ class CADEEasyUI(ComfyAdaptiveDetailEnhancer25):
                    model, positive, negative, vae, latent,
                    seed, steps, cfg, denoise, sampler_name, scheduler,
                    clipseg_text="", reference_image=None, clip_vision=None, custom=False):
-        lat, img, _s, _c, _d, mask = super().apply_cade2(
+        lat, img, _s, _c, _d, mask, selected_seed = super().apply_cade2(
             model, vae, positive, negative, latent,
             int(seed), int(steps), float(cfg), float(denoise),
             str(sampler_name), str(scheduler), 0.0,
@@ -3478,11 +3478,11 @@ class CADEEasyUI(ComfyAdaptiveDetailEnhancer25):
             reference_image=reference_image,
             clip_vision=clip_vision,
         )
-        return lat, img, mask
+        return lat, img, mask, selected_seed
 
         # Show simpler outputs in Easy variant
-        RETURN_TYPES = ("LATENT", "IMAGE", "IMAGE")
-        RETURN_NAMES = ("LATENT", "IMAGE", "mask_preview")
+        RETURN_TYPES = ("LATENT", "IMAGE", "IMAGE", "INT")
+        RETURN_NAMES = ("LATENT", "IMAGE", "mask_preview", "selected_seed")
         FUNCTION = "apply_easy"
 
         def apply_easy(self,
@@ -3490,13 +3490,13 @@ class CADEEasyUI(ComfyAdaptiveDetailEnhancer25):
                        model, positive, negative, vae, latent,
                        seed, steps, cfg, denoise, sampler_name, scheduler,
                        clipseg_text=""):
-            lat, img, _s, _c, _d, mask = super().apply_cade2(
+            lat, img, _s, _c, _d, mask, selected_seed = super().apply_cade2(
                 model, vae, positive, negative, latent,
                 int(seed), int(steps), float(cfg), float(denoise),
                 str(sampler_name), str(scheduler), 0.0,
                 preset_step=str(preset_step), custom_override=bool(custom), clipseg_text=str(clipseg_text),
             )
-            return lat, img, mask
+            return lat, img, mask, selected_seed
 
 
 

mod/easy/mg_supersimple_easy.py

@@ -13,7 +13,7 @@ control_net: ControlNet module for CF (required)
 reference_image/clip_vision: forwarded into CADE (optional)
 Outputs:
 
-(LATENT, IMAGE) from the final executed step
+(LATENT, IMAGE, selected_seed) from the final executed step
 """
 
 
@@ -62,8 +62,8 @@ class MG_SuperSimple:
             },
         }
 
-    RETURN_TYPES = ("LATENT", "IMAGE")
-    RETURN_NAMES = ("LATENT", "IMAGE")
+    RETURN_TYPES = ("LATENT", "IMAGE", "INT")
+    RETURN_NAMES = ("LATENT", "IMAGE", "selected_seed")
     FUNCTION = "run"
 
     def _cade(self,
@@ -75,7 +75,7 @@ class MG_SuperSimple:
               clipseg_text: str,
               reference_image=None, clip_vision=None):
         # CADE core call mirrors CADEEasyUI -> apply_cade2
-        lat, img, _s, _c, _d, _mask = _CADE().apply_cade2(
+        lat, img, _s, _c, _d, _mask, selected_seed = _CADE().apply_cade2(
             model, vae, positive, negative, latent,
             int(seed), int(steps), float(cfg), float(denoise),
             str(sampler_name), str(scheduler), 0.0,
@@ -83,7 +83,7 @@ class MG_SuperSimple:
             clipseg_text=str(clipseg_text),
             reference_image=reference_image, clip_vision=clip_vision,
         )
-        return lat, img
+        return lat, img, selected_seed
 
     def _cf(self,
             preset_step: str,
@@ -112,11 +112,12 @@ class MG_SuperSimple:
         cur_image = None
         cur_pos = positive
         cur_neg = negative
+        selected_seed = int(seed)
 
         try:
             # Step 1: CADE with Step 1 preset, denoise forced to 1.0
             denoise_step1 = 1.0
-            lat1, img1 = self._cade(
+            lat1, img1, selected_seed = self._cade(
                 preset_step="Step 1",
                 custom_override=bool(custom),
                 model=model, vae=vae, positive=cur_pos, negative=cur_neg, latent=cur_latent,
@@ -141,7 +142,7 @@ class MG_SuperSimple:
                 # If no external reference_image is provided, use the previous step image
                 # so that reference_clean / CLIP-Vision gating can take effect.
                 ref_img = reference_image if (reference_image is not None) else cur_image
-                lat_i, img_i = self._cade(
+                lat_i, img_i, selected_seed = self._cade(
                     preset_step=f"Step {i}",
                     custom_override=bool(custom),
                     model=model, vae=vae, positive=cur_pos, negative=cur_neg, latent=cur_latent,
@@ -151,7 +152,7 @@ class MG_SuperSimple:
                     reference_image=ref_img, clip_vision=clip_vision,
                 )
                 cur_latent, cur_image = lat_i, img_i
-            return (cur_latent, cur_image)
+            return (cur_latent, cur_image, selected_seed)
         finally:
             # try to free memory on cancel or normal exit
             try:

tests/test_smartseed_outputs.py

@@ -0,0 +1,77 @@
+import ast
+from pathlib import Path
+import unittest
+
+
+ROOT = Path(__file__).resolve().parents[1]
+
+
+def _module_tree(relative_path: str) -> ast.Module:
+    return ast.parse((ROOT / relative_path).read_text(encoding="utf-8"))
+
+
+def _class_node(tree: ast.Module, class_name: str) -> ast.ClassDef:
+    for node in tree.body:
+        if isinstance(node, ast.ClassDef) and node.name == class_name:
+            return node
+    raise AssertionError(f"class {class_name} not found")
+
+
+def _assigned_tuple(class_node: ast.ClassDef, name: str) -> tuple[str, ...]:
+    for node in class_node.body:
+        if (
+            isinstance(node, ast.Assign)
+            and any(isinstance(target, ast.Name) and target.id == name for target in node.targets)
+            and isinstance(node.value, ast.Tuple)
+        ):
+            return tuple(
+                item.value
+                for item in node.value.elts
+                if isinstance(item, ast.Constant) and isinstance(item.value, str)
+            )
+    raise AssertionError(f"{name} tuple not found on {class_node.name}")
+
+
+class SmartSeedOutputTests(unittest.TestCase):
+    def test_easy_cade_exposes_selected_seed_without_reordering_existing_outputs(self):
+        tree = _module_tree("mod/easy/mg_cade25_easy.py")
+        cade_easy = _class_node(tree, "CADEEasyUI")
+
+        self.assertEqual(
+            _assigned_tuple(cade_easy, "RETURN_NAMES"),
+            ("LATENT", "IMAGE", "mask_preview", "selected_seed"),
+        )
+        self.assertEqual(
+            _assigned_tuple(cade_easy, "RETURN_TYPES"),
+            ("LATENT", "IMAGE", "IMAGE", "INT"),
+        )
+
+    def test_base_cade_returns_selected_seed_for_advanced_wiring(self):
+        tree = _module_tree("mod/easy/mg_cade25_easy.py")
+        base_cade = _class_node(tree, "ComfyAdaptiveDetailEnhancer25")
+
+        self.assertEqual(
+            _assigned_tuple(base_cade, "RETURN_NAMES"),
+            ("LATENT", "IMAGE", "steps", "cfg", "denoise", "mask_preview", "selected_seed"),
+        )
+        self.assertEqual(
+            _assigned_tuple(base_cade, "RETURN_TYPES"),
+            ("LATENT", "IMAGE", "INT", "FLOAT", "FLOAT", "IMAGE", "INT"),
+        )
+
+    def test_supersimple_exposes_final_step_selected_seed(self):
+        tree = _module_tree("mod/easy/mg_supersimple_easy.py")
+        super_simple = _class_node(tree, "MG_SuperSimple")
+
+        self.assertEqual(
+            _assigned_tuple(super_simple, "RETURN_NAMES"),
+            ("LATENT", "IMAGE", "selected_seed"),
+        )
+        self.assertEqual(
+            _assigned_tuple(super_simple, "RETURN_TYPES"),
+            ("LATENT", "IMAGE", "INT"),
+        )
+
+
+if __name__ == "__main__":
+    unittest.main()