Add handoff context frames

README.md

@@ -38,6 +38,7 @@
   - The local Dumas prompt-engineering fork has been removed from this node. Long Videos now wraps the upstream sampler/engine directly so it can track the source project again.
   - Upstream compatibility keys `H3LongVideos`, `H3LongVideosFL2VA`, `H3LongVideosV1`, and `H3LongVideosREF2VA` are also registered to the same class.
   - The old Dumas browser widget grouping script is disabled for this node because it targeted controls that no longer exist on the upstream sampler.
+  - `handoff_frames` extends the upstream last-frame handoff: `1` keeps the current single keyframe behavior; higher values keep that final-frame keyframe and add earlier tail frames from the previous shot as claimed reference context for the next beat.
   - Upstream license text is included in [`H3_LONGVIDEOS_UPSTREAM_LICENSE.txt`](./H3_LONGVIDEOS_UPSTREAM_LICENSE.txt).
 
 - `Dumas H3 Latent Upscale Params`
@@ -48,7 +49,9 @@
 - `Dumas H3 Beat Prompt`
   - Inputs: authored through the custom front-end beat editor
   - Output: `prompt`
-  - Builds one H3 prompt block per beat, with quick controls for per-shot timing, continuity, ref behavior, anchor additions, soundscape, and music while staying compatible with direct text editing.
+  - Builds an upstream-compatible Long Videos prompt: optional scene paragraph, optional character sheet, then one blank-line-separated textbox per beat.
+  - Per-beat helpers only emit upstream-supported state directives: `remove:` / `removed:` / `off:` and `add:` / `wear:` / `wearing:`.
+  - Old Dumas-only beat directives such as `seconds:`, `continuity:`, `ref_mode:`, `ref_noise_aug:`, `anchor_add:`, `soundscape:`, and `music:` are stripped from the generated prompt so they are not sent to the upstream node as visible text.
 
 - `Dumas H3 Prompt Curator`
   - Inputs: `action_prompt`, `anatomy_guard`, `subject_count_guard`, optional `anchor`, optional `soundscape`, optional `bgm`, optional `ref_1` through `ref_9`

dumas_h3_beat_prompt.py

@@ -2,11 +2,46 @@ import json
 
 
 _DEFAULT_BEAT = "Describe this beat."
-_DEFAULT_STATE = {"beats": [{"text": _DEFAULT_BEAT}]}
+_DEFAULT_STATE = {"scene": "", "character_sheet": "", "beats": [{"text": _DEFAULT_BEAT}]}
+_LEGACY_DIRECTIVE_PREFIXES = (
+    "seconds",
+    "duration",
+    "continuity",
+    "ref_mode",
+    "ref_noise_aug",
+    "anchor_add",
+    "overall_soundscape",
+    "soundscape",
+    "non_diegetic_music",
+    "music",
+    "wardrobe",
+    "enter",
+    "exit",
+)
 
 
 def _clone_default_state():
-    return {"beats": [{"text": _DEFAULT_BEAT}]}
+    return {
+        "scene": "",
+        "character_sheet": "",
+        "beats": [{"text": _DEFAULT_BEAT}],
+    }
+
+
+def _strip_legacy_directives(text):
+    """Remove directives from the abandoned Dumas Long Videos fork.
+
+    The upstream Long Videos node sends unknown field labels to the model as text,
+    so this builder strips the old managed controls rather than emitting prompts
+    that ask H3 to draw labels such as "seconds:" or "music:" in the frame.
+    """
+    kept = []
+    for line in str(text or "").splitlines():
+        lowered = line.strip().lower()
+        if any(lowered.startswith(f"{name}:") for name in _LEGACY_DIRECTIVE_PREFIXES):
+            continue
+        kept.append(line)
+    return "\n".join(kept).strip()
 
 
 def _parse_beat_prompt_state(value):
@@ -21,6 +56,8 @@ def _parse_beat_prompt_state(value):
         except Exception:
             return _clone_default_state()
 
+    scene = str(raw.get("scene") or "")
+    character_sheet = str(raw.get("character_sheet") or "")
     beats = []
     for item in list(raw.get("beats") or []):
         if isinstance(item, dict):
@@ -30,15 +67,25 @@ def _parse_beat_prompt_state(value):
         beats.append({"text": text})
 
     if not beats:
-        return _clone_default_state()
-    return {"beats": beats}
+        beats = [{"text": _DEFAULT_BEAT}]
+    return {
+        "scene": scene,
+        "character_sheet": character_sheet,
+        "beats": beats,
+    }
 
 
 def _assemble_beat_prompt(state):
     parsed = _parse_beat_prompt_state(state)
     chunks = []
+    scene = str(parsed.get("scene") or "").strip()
+    if scene:
+        chunks.append(scene)
+    character_sheet = str(parsed.get("character_sheet") or "").strip()
+    if character_sheet:
+        chunks.append(character_sheet)
     for beat in parsed["beats"]:
-        text = str(beat.get("text") or "").strip()
+        text = _strip_legacy_directives(beat.get("text") or "")
         if text:
             chunks.append(text)
     return "\n\n".join(chunks)
@@ -46,9 +93,9 @@ def _assemble_beat_prompt(state):
 
 class DumasH3BeatPromptNode:
     DESCRIPTION = (
-        "Build a MiniMax H3 prompt from one textbox per beat, with a front-end beat "
-        "editor that can append directive examples and expose per-shot controls for "
-        "timing, continuity, ref behavior, anchor additions, soundscape, and music."
+        "Build an upstream MiniMax H3 Long Videos prompt: optional scene paragraph, "
+        "optional character sheet, then one blank-line-separated textbox per beat. "
+        "Per-beat helpers only emit directives the upstream node understands."
     )
     RETURN_TYPES = ("STRING",)
     RETURN_NAMES = ("prompt",)

dumas_h3_longvideos_upstream.py

@@ -5831,6 +5831,24 @@ def handoff_claim(n):
             f"anybody new.")
 
 
+def handoff_context_claim(first, last):
+    """Claim previous-shot tail frames carried as reference context.
+
+    These are not keyframes. They are context stills that show the room, camera
+    path, and motion immediately before the hard handoff keyframe. They must be
+    named in the prompt because every reference image the VLM sees needs a job:
+    unnamed pictures are free to become extra subjects.
+    """
+    first, last = int(first), int(last)
+    if last <= first:
+        return (f" <Picture {first}> is a continuity context frame from immediately "
+                f"before this shot: same place, same camera path, same people, no "
+                f"new subject.")
+    return (f" <Picture {first}> through <Picture {last}> are continuity context "
+            f"frames from immediately before this shot: same place, same camera "
+            f"path, same people, no new subjects.")
+
+
 def room_claim(n, present, joining):
     """Claim a handoff carried as a reference because somebody NEW is in the shot.
 
@@ -7062,6 +7080,7 @@ _WIDGET_RANGE = {
     "pace": (1.0, 0.25, 2.0, float),
     "ambient_level": (0.25, 0.0, 1.0, float),
     "foley_level": (0.35, 0.0, 1.0, float),
+    "handoff_frames": (1, 1, MAX_FRAMES, int),
 }
 
 
@@ -7582,6 +7601,18 @@ class H3LongVideos:
                                "It is synthesis, not a recording: a click, a rattle, "
                                "a rustle, in the right place. Nothing vocal is ever "
                                "built. 0 turns it off; needs auto_sound on."}),
+                # APPENDED. Saved workflows restore widget values by position.
+                "handoff_frames": ("INT", {"default": 1, "min": 1, "max": MAX_FRAMES,
+                    "tooltip": "How many frames from the previous shot are used to "
+                               "condition the next one. 1 is the upstream default: "
+                               "the previous final frame becomes the next shot's "
+                               "keyframe. Values above 1 keep that final-frame "
+                               "keyframe and append the earlier tail frames as "
+                               "claimed reference context, so the next beat can see "
+                               "more of the incoming motion and room continuity. "
+                               "The extra frames are references, not extra "
+                               "keyframes, so seam trimming still removes only the "
+                               "duplicated opening frame."}),
             },
         }
 
@@ -7608,6 +7639,7 @@ class H3LongVideos:
             character_guard=True, pace=1.0, auto_sound=True, hold_scene_state=True,
             mouths_shut_when_no_line=True, hold_gaze=True,
             ambient_audio=None, ambient_level=0.25, foley_level=0.35,
+            handoff_frames=1,
             **_removed):
         # **_removed: a workflow saved with the old `save_defaults` widget still sends
         # it. Swallowed rather than raising, so an existing workflow keeps loading.
@@ -7632,7 +7664,8 @@ class H3LongVideos:
             shift_video=shift_video, shift_audio=shift_audio,
             ref_noise_aug=ref_noise_aug, latent_upscale_scale=latent_upscale_scale,
             upscale_target_short_edge=upscale_target_short_edge,
-            upscale_batch=upscale_batch, pace=pace))
+            upscale_batch=upscale_batch, pace=pace,
+            handoff_frames=handoff_frames))
         megapixels, shot_seconds = _fixed["megapixels"], _fixed["shot_seconds"]
         steps, cfg = _fixed["steps"], _fixed["cfg"]
         shift_video, shift_audio = _fixed["shift_video"], _fixed["shift_audio"]
@@ -7640,6 +7673,7 @@ class H3LongVideos:
         latent_upscale_scale = _fixed["latent_upscale_scale"]
         upscale_target_short_edge = _fixed["upscale_target_short_edge"]
         upscale_batch, pace = _fixed["upscale_batch"], _fixed["pace"]
+        handoff_frames = _fixed["handoff_frames"]
         notes.extend(_fixnotes)
         # <Picture N> means ref_image_N, the socket. Everything downstream works on
         # the packed roster instead, so translate once, here, before anything has
@@ -10234,6 +10268,7 @@ class H3LongVideos:
             negative = clip.encode_from_tokens_scheduled(clip.tokenize(""))
 
         handoff = first_frame
+        handoff_context = None
         # Where the time actually goes. Sampling and decode trade off against each
         # other -- latent_upscale buys cheaper sampling and pays for it at decode,
         # and which side wins depends on `steps`. Reported so the trade is a
@@ -10365,6 +10400,20 @@ class H3LongVideos:
             # here rather than inside build_conditioning because the claim is text,
             # and the text is assembled up here.
             _shot_refs = list(shot_refs_all[i]) + _extra
+            _ctx_refs = []
+            _keyframe_ok = ref_noise_aug is None or float(ref_noise_aug) >= KEYFRAME_SAFE_AUG
+            if (handoff_frames > 1 and shot_handoff is not None and handoff_context is not None
+                    and not _handoff_ref and _keyframe_ok):
+                try:
+                    _ctx_refs = [handoff_context[j:j + 1]
+                                 for j in range(int(handoff_context.shape[0]))]
+                except Exception:
+                    _ctx_refs = []
+                if _ctx_refs:
+                    _first = len(_shot_refs) + 1
+                    _last = _first + len(_ctx_refs) - 1
+                    shot_prompt = shot_prompt + handoff_context_claim(_first, _last)
+                    _shot_refs.extend(_ctx_refs)
             if _handoff_ref:
                 # Carried for the ROOM, with somebody new in the shot -- so the
                 # standing claim is exactly wrong here ("joined by anybody new") and
@@ -10380,7 +10429,7 @@ class H3LongVideos:
             sent_text[i] = shot_prompt
             cond, latent, fc, demoted = build_conditioning(
                 clip, vae, audio_vae, shot_prompt, w, h, lens[i],
-                handoff=shot_handoff, refs=list(shot_refs_all[i]) + _extra,
+                handoff=shot_handoff, refs=_shot_refs,
                 ref_noise_aug=ref_noise_aug, silent=silent,
                 handoff_as_ref=_handoff_ref)
             if demoted and not _aug_warned:
@@ -10457,6 +10506,16 @@ class H3LongVideos:
             # 0..1, and feeding that back in to be re-encoded every boundary is a
             # drift that accumulates rather than cancels.
             handoff = hand_src[-1:].detach().clamp(0.0, 1.0).to("cpu", copy=True)
+            handoff_context = None
+            if handoff_frames > 1:
+                try:
+                    available = max(0, int(hand_src.shape[0]) - 1)
+                    want = min(max(0, int(handoff_frames) - 1), available)
+                    if want > 0:
+                        handoff_context = hand_src[-(want + 1):-1].detach().clamp(
+                            0.0, 1.0).to("cpu", copy=True)
+                except Exception:
+                    handoff_context = None
             # Keep a frame for the shot they come back on -- but ONLY from a shot that
             # was theirs alone.
             #

js/h3_beat_prompt/index.js

@@ -8,32 +8,15 @@ const DEFAULT_W = 520;
 const DEFAULT_H = 340;
 const DEFAULT_BEAT = "Describe this beat.";
 const STATE_PROPERTY = "dumas_h3_beat_prompt_state";
-const CONTINUITY_OPTIONS = ["", "soft carry", "hard cut", "keyframe carry", "handoff ref"];
-const REF_MODE_OPTIONS = ["", "auto ref2v", "where tagged", "first shot", "every shot", "every shot + handoff ref"];
 const MANAGED_DIRECTIVES = {
-  seconds: ["seconds", "duration"],
-  continuity: ["continuity"],
-  ref_mode: ["ref_mode"],
-  ref_noise_aug: ["ref_noise_aug"],
-  anchor_add: ["anchor_add"],
-  overall_soundscape: ["overall_soundscape", "soundscape"],
-  non_diegetic_music: ["non_diegetic_music", "music"],
+  remove: ["remove", "removed", "off"],
+  add: ["add", "wear", "wearing"],
 };
 const DIRECTIVE_EXAMPLES = [
-  ["wardrobe set", "wardrobe: Maya = grey shorts, red jacket"],
-  ["wardrobe add", "wardrobe: Maya += red jacket"],
-  ["wardrobe remove", "wardrobe: Maya -= red jacket"],
-  ["seconds", "seconds: 8"],
-  ["exit", "exit: Maya"],
-  ["enter", "enter: Jon"],
-  ["continuity", "continuity: hard cut"],
-  ["ref_mode", "ref_mode: every shot"],
-  ["ref_noise_aug", "ref_noise_aug: 0.92"],
-  ["anchor_add", "anchor_add: harsh sodium-vapor spill, wet pavement, long-lens compression"],
-  ["overall_soundscape", "overall_soundscape: soft rain, distant traffic"],
-  ["non_diegetic_music", "non_diegetic_music: tense analog synth pulse"],
-  ["soundscape", "soundscape: fluorescent room tone, faint HVAC hum"],
-  ["music", "music: low ominous cello and sparse percussion"],
+  ["remove", "remove: red jacket"],
+  ["off", "off: steel collar"],
+  ["add", "add: white shirt underneath"],
+  ["wearing", "wearing: black coat"],
 ];
 
 function injectCSS() {
@@ -183,7 +166,7 @@ function injectCSS() {
 }
 
 function defaultState() {
-  return { beats: [{ text: DEFAULT_BEAT }] };
+  return { scene: "", character_sheet: "", beats: [{ text: DEFAULT_BEAT }] };
 }
 
 function normalizeState(value) {
@@ -200,7 +183,11 @@ function normalizeState(value) {
   const normalized = beats.map((beat) => ({
     text: typeof beat?.text === "string" ? beat.text : String(beat?.text || ""),
   }));
-  return normalized.length ? { beats: normalized } : defaultState();
+  return {
+    scene: typeof parsed.scene === "string" ? parsed.scene : String(parsed.scene || ""),
+    character_sheet: typeof parsed.character_sheet === "string" ? parsed.character_sheet : String(parsed.character_sheet || ""),
+    beats: normalized.length ? normalized : [{ text: DEFAULT_BEAT }],
+  };
 }
 
 function readState(node) {
@@ -321,6 +308,51 @@ function renderUI(node) {
   node._dh3bpRenderedState = JSON.stringify(state);
   ui.list.innerHTML = "";
 
+  const buildTopTextarea = ({ labelText, placeholder, value, onInput }) => {
+    const card = document.createElement("div");
+    card.className = "dh3bp-beat";
+
+    const label = document.createElement("div");
+    label.className = "dh3bp-label";
+    label.textContent = labelText;
+
+    const textarea = document.createElement("textarea");
+    textarea.className = "dh3bp-text";
+    textarea.placeholder = placeholder;
+    textarea.value = value || "";
+    textarea.addEventListener("input", () => {
+      onInput(textarea.value);
+      updateTextareaHeight(textarea);
+    });
+    textarea.addEventListener("keydown", stopCanvasKeyboard);
+
+    card.append(label, textarea);
+    updateTextareaHeight(textarea);
+    return card;
+  };
+
+  ui.list.appendChild(buildTopTextarea({
+    labelText: "Scene paragraph",
+    placeholder: "Optional. Persistent location, lighting, camera, tone. Leave empty if you wire the Long Videos anchor input.",
+    value: state.scene,
+    onInput: (value) => {
+      const next = readState(node);
+      next.scene = value;
+      writeState(node, next);
+    },
+  }));
+
+  ui.list.appendChild(buildTopTextarea({
+    labelText: "Character sheet",
+    placeholder: "Optional. One character per line, e.g. Maya: 27, she, silver hair, red jacket, the woman in <Picture 1>.",
+    value: state.character_sheet,
+    onInput: (value) => {
+      const next = readState(node);
+      next.character_sheet = value;
+      writeState(node, next);
+    },
+  }));
+
   state.beats.forEach((beat, index) => {
     const card = document.createElement("div");
     card.className = "dh3bp-beat";
@@ -379,85 +411,27 @@ function renderUI(node) {
       return wrap;
     };
 
-    const secondsInput = document.createElement("input");
-    secondsInput.className = "dh3bp-input";
-    secondsInput.type = "text";
-    secondsInput.placeholder = "8";
-    secondsInput.value = readDirectiveValue(beat.text, MANAGED_DIRECTIVES.seconds);
-    secondsInput.addEventListener("input", () => {
-      applyTextUpdate(setDirectiveValue(textarea.value, "seconds", MANAGED_DIRECTIVES.seconds, secondsInput.value));
+    const removeInput = document.createElement("input");
+    removeInput.className = "dh3bp-input";
+    removeInput.type = "text";
+    removeInput.placeholder = "red jacket";
+    removeInput.value = readDirectiveValue(beat.text, MANAGED_DIRECTIVES.remove);
+    removeInput.addEventListener("input", () => {
+      applyTextUpdate(setDirectiveValue(textarea.value, "remove", MANAGED_DIRECTIVES.remove, removeInput.value));
     });
 
-    const continuitySelect = document.createElement("select");
-    continuitySelect.className = "dh3bp-select";
-    CONTINUITY_OPTIONS.forEach((value) => {
-      const option = document.createElement("option");
-      option.value = value;
-      option.textContent = value || "Default";
-      continuitySelect.appendChild(option);
-    });
-    continuitySelect.value = readDirectiveValue(beat.text, MANAGED_DIRECTIVES.continuity);
-    continuitySelect.addEventListener("change", () => {
-      applyTextUpdate(setDirectiveValue(textarea.value, "continuity", MANAGED_DIRECTIVES.continuity, continuitySelect.value));
-    });
-
-    const refModeSelect = document.createElement("select");
-    refModeSelect.className = "dh3bp-select";
-    REF_MODE_OPTIONS.forEach((value) => {
-      const option = document.createElement("option");
-      option.value = value;
-      option.textContent = value || "Global";
-      refModeSelect.appendChild(option);
-    });
-    refModeSelect.value = readDirectiveValue(beat.text, MANAGED_DIRECTIVES.ref_mode);
-    refModeSelect.addEventListener("change", () => {
-      applyTextUpdate(setDirectiveValue(textarea.value, "ref_mode", MANAGED_DIRECTIVES.ref_mode, refModeSelect.value));
-    });
-
-    const refNoiseInput = document.createElement("input");
-    refNoiseInput.className = "dh3bp-input";
-    refNoiseInput.type = "text";
-    refNoiseInput.placeholder = "0.95";
-    refNoiseInput.value = readDirectiveValue(beat.text, MANAGED_DIRECTIVES.ref_noise_aug);
-    refNoiseInput.addEventListener("input", () => {
-      applyTextUpdate(setDirectiveValue(textarea.value, "ref_noise_aug", MANAGED_DIRECTIVES.ref_noise_aug, refNoiseInput.value));
-    });
-
-    const anchorInput = document.createElement("input");
-    anchorInput.className = "dh3bp-input";
-    anchorInput.type = "text";
-    anchorInput.placeholder = "extra per-shot style treatment";
-    anchorInput.value = readDirectiveValue(beat.text, MANAGED_DIRECTIVES.anchor_add);
-    anchorInput.addEventListener("input", () => {
-      applyTextUpdate(setDirectiveValue(textarea.value, "anchor_add", MANAGED_DIRECTIVES.anchor_add, anchorInput.value));
-    });
-
-    const soundscapeInput = document.createElement("input");
-    soundscapeInput.className = "dh3bp-input";
-    soundscapeInput.type = "text";
-    soundscapeInput.placeholder = "faint traffic, loose sign rattle";
-    soundscapeInput.value = readDirectiveValue(beat.text, MANAGED_DIRECTIVES.overall_soundscape);
-    soundscapeInput.addEventListener("input", () => {
-      applyTextUpdate(setDirectiveValue(textarea.value, "overall_soundscape", MANAGED_DIRECTIVES.overall_soundscape, soundscapeInput.value));
-    });
-
-    const musicInput = document.createElement("input");
-    musicInput.className = "dh3bp-input";
-    musicInput.type = "text";
-    musicInput.placeholder = "low pulsing synth tension";
-    musicInput.value = readDirectiveValue(beat.text, MANAGED_DIRECTIVES.non_diegetic_music);
-    musicInput.addEventListener("input", () => {
-      applyTextUpdate(setDirectiveValue(textarea.value, "non_diegetic_music", MANAGED_DIRECTIVES.non_diegetic_music, musicInput.value));
+    const addInput = document.createElement("input");
+    addInput.className = "dh3bp-input";
+    addInput.type = "text";
+    addInput.placeholder = "white shirt underneath";
+    addInput.value = readDirectiveValue(beat.text, MANAGED_DIRECTIVES.add);
+    addInput.addEventListener("input", () => {
+      applyTextUpdate(setDirectiveValue(textarea.value, "add", MANAGED_DIRECTIVES.add, addInput.value));
     });
 
     controls.append(
-      buildField({ labelText: "Seconds", input: secondsInput }),
-      buildField({ labelText: "Continuity", input: continuitySelect }),
-      buildField({ labelText: "Ref Mode", input: refModeSelect }),
-      buildField({ labelText: "Ref Noise Aug", input: refNoiseInput }),
-      buildField({ labelText: "Anchor Add", className: "dh3bp-control-wide", input: anchorInput }),
-      buildField({ labelText: "Shot Soundscape", className: "dh3bp-control-wide", input: soundscapeInput }),
-      buildField({ labelText: "Shot Music", className: "dh3bp-control-wide", input: musicInput }),
+      buildField({ labelText: "Remove from memory", input: removeInput }),
+      buildField({ labelText: "Add to memory", input: addInput }),
     );
 
     const directives = document.createElement("div");
@@ -501,7 +475,7 @@ function setupNode(node) {
   title.textContent = "Beat Prompt Builder";
   const subtitle = document.createElement("div");
   subtitle.className = "dh3bp-subtitle";
-  subtitle.textContent = "One textbox per H3 beat, plus per-shot controls for timing, ref behavior, continuity, anchor adds, and audio directives.";
+  subtitle.textContent = "Upstream Long Videos format: optional scene, optional character sheet, then one blank-line-separated beat per shot.";
   titleWrap.append(title, subtitle);
 
   const addButton = document.createElement("button");

tests/test_dumas_h3_beat_prompt.py

@@ -11,35 +11,62 @@ class DumasH3BeatPromptTests(unittest.TestCase):
         state = self.module._parse_beat_prompt_state("not json")
         self.assertEqual(
             state,
-            {"beats": [{"text": "Describe this beat."}]},
+            {
+                "scene": "",
+                "character_sheet": "",
+                "beats": [{"text": "Describe this beat."}],
+            },
         )
 
-    def test_assemble_prompt_joins_beats_with_blank_lines(self):
+    def test_assemble_prompt_outputs_upstream_sections(self):
         prompt = self.module._assemble_beat_prompt(
             {
+                "scene": "A rainy kitchen at night.",
+                "character_sheet": "Maya: 27, she, red jacket, silver hair.",
                 "beats": [
-                    {"text": "A woman enters the room."},
-                    {"text": "wardrobe: Maya = red jacket\nShe sits at the table."},
+                    {"text": "Maya enters the room."},
+                    {"text": "remove: red jacket\nadd: white shirt underneath\nShe sits at the table."},
                     {"text": "  "},
-                    {"text": "music: low synth pulse"},
                 ]
             }
         )
         self.assertEqual(
             prompt,
             (
-                "A woman enters the room.\n\n"
-                "wardrobe: Maya = red jacket\nShe sits at the table.\n\n"
-                "music: low synth pulse"
+                "A rainy kitchen at night.\n\n"
+                "Maya: 27, she, red jacket, silver hair.\n\n"
+                "Maya enters the room.\n\n"
+                "remove: red jacket\nadd: white shirt underneath\nShe sits at the table."
             ),
         )
 
+    def test_assemble_prompt_strips_old_dumas_directives(self):
+        prompt = self.module._assemble_beat_prompt(
+            {
+                "beats": [
+                    {
+                        "text": (
+                            "seconds: 8\n"
+                            "continuity: hard cut\n"
+                            "ref_mode: every shot\n"
+                            "soundscape: soft rain\n"
+                            "music: low synth\n"
+                            "Maya opens the cupboard.\n"
+                            "remove: red jacket"
+                        )
+                    },
+                ]
+            }
+        )
+
+        self.assertEqual(prompt, "Maya opens the cupboard.\nremove: red jacket")
+
     def test_node_build_prompt_uses_hidden_state(self):
         node = self.module.DumasH3BeatPromptNode()
         result = node.build_prompt(
-            '{"beats":[{"text":"Beat one"},{"text":"Beat two"}]}'
+            '{"scene":"Scene","character_sheet":"Maya: 27, she","beats":[{"text":"Beat one"},{"text":"Beat two"}]}'
         )
-        self.assertEqual(result, ("Beat one\n\nBeat two",))
+        self.assertEqual(result, ("Scene\n\nMaya: 27, she\n\nBeat one\n\nBeat two",))
 
 
 if __name__ == "__main__":

tests/test_dumas_h3_longvideos.py

@@ -112,8 +112,17 @@ class DumasH3LongVideosUpstreamWrapperTests(unittest.TestCase):
         self.assertIn("first_frame", schema["optional"])
         self.assertIn("ref_image_1", schema["optional"])
         self.assertIn("latent_upscale", schema["optional"])
+        self.assertIn("handoff_frames", schema["optional"])
+        self.assertEqual(schema["optional"]["handoff_frames"][1]["default"], 1)
         self.assertEqual(node_cls.RETURN_NAMES[0:4], ("images", "audio", "info", "script"))
 
+    def test_handoff_context_claim_names_reference_range(self):
+        upstream = importlib.import_module("dumas_h3_longvideos_upstream")
+
+        self.assertIn("<Picture 2> through <Picture 22>", upstream.handoff_context_claim(2, 22))
+        self.assertIn("no new subjects", upstream.handoff_context_claim(2, 22))
+        self.assertIn("<Picture 5>", upstream.handoff_context_claim(5, 5))
+
 
 class _NullContext:
     def __enter__(self):