@remixmate/cli 0.12.5 → 0.12.7

package/dist/manifest.json

@@ -1,7 +1,7 @@
 {
   "schemaVersion": 1,
-  "version": "0.12.5",
-  "generatedAt": "2026-10-18T12:54:02.484Z",
+  "version": "0.12.7",
+  "generatedAt": "2026-10-20T11:33:13.520Z",
   "skills": [
     {
       "id": "doc-parse",

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@remixmate/cli",
-  "version": "0.12.5",
+  "version": "0.12.7",
   "description": "AI media generation skills for Claude Code / Codex — 12 skills covering image, video, voice, digital human, web screenshot, web recording, script, template registry, rendering, Jianying export, and video deconstruction.",
   "type": "module",
   "main": "dist/index.js",

package/skills/gen-script/SKILL.md

@@ -78,7 +78,7 @@ The intent is to avoid the failure mode where "the generic DSL looks compatible
 4. Combine that with the template's `slotMapping`, `requiredProps`, `optionalProps`, `propExtractors`, `assetRequirements`, `supportedAspectRatios`, `supportedDurations`, `constraints`, and `scenePatterns`.
 5. Generate the DSL in the template's native shape — not the generic DSL shape.
 6. Make sure the DSL explicitly contains every template-specific field, e.g. `templateData.words`, `customPayload.slideId`, `visuals.avatar.assetRef`.
-7. After generation, run the schema check and show the script summary to the user for confirmation.
+7. After generation, run `gen_script.py --validate -i <dsl.json>` and fix every error it reports, then show the script summary to the user for confirmation. It checks the template contract, not just the DSL shape: undeclared `customPayload` field names, values outside a declared `enum` (e.g. `layout: "bogus"`, `style.preset: "nwes"`), and every `assetRef` / `*AssetRef` (`logoAssetRef`, `avatarAssetRef` …) pointing at an asset that `assets[]` does not declare. These are the same checks render-video runs before spending credits; catching them here saves a confirmed script from being rejected later.
 
 ### Document-driven templates: pass the source URL, don't author the payload
 
@@ -123,7 +123,7 @@ Picking the layout is only half of it. Each layout reads its own field names out
 
 This has shipped broken videos: a 7-scene deck guessed generic names (`title`, `description`, `ctaText`) where the chosen layouts read different ones. Five of seven scenes rendered as empty cards and placeholder copy — after paying for TTS and the render. Which names each layout reads is the template's fact; the template's `llmHint` calls out the usual traps.
 
-So: **copy the field names out of `slideSchemas[slideId]`.** The DSL validator now rejects unknown field names before any asset is generated and tells you the declared ones, but that check only fires for templates that ship `slideSchemas` — reading the contract is still the primary move, not the fallback.
+So: **copy the field names out of `slideSchemas[slideId]`.** The DSL validator (`gen_script.py --validate`, and again in render-video) rejects unknown field names and out-of-enum values before any asset is generated and tells you the declared ones — per layout from `slideSchemas` for multi-layout templates, from `customPayloadSchema` for every other template. Reading the contract is still the primary move, not the fallback: the validator checks names and enums, not whether the content fits the layout.
 
 ### On-screen text: the rules live in the template, not here
 
@@ -283,6 +283,12 @@ Without this flag the user's images are **ignored entirely** and every scene get
 AI-generated picture. That failure is quiet: the video renders fine, it just isn't made of
 the material the user handed you.
 
+**One shared background.** A template that declares `capabilities.backgroundStrategy: "shared"`
+(e.g. an audiogram whose whole video is one "show poster") gets a skeleton with a single
+`img-background` asset that every scene references, instead of one image per scene. Keep it
+that way; switch to another image only where the template's `llmHint` says you may (such as a
+new chapter). `--scene-images` still maps the user's own images one per scene.
+
 ### The mapping rule (positional, then AI fills the rest)
 
 Each entry is one scene's visual, and it is either a **URL** (use that existing asset) or the
@@ -473,7 +479,7 @@ python3 <SkillDir>/scripts/gen_script.py \
 | `--scenes` | Scene count. | auto-planned |
 | `--voice-id` | Narration voice id. | resolved from template + language fallback |
 | `--speed` | Narration speech rate (0.5–2.0), written to `global.narration.speed` and applied when `prepare_video_assets` calls gen-voice. Pass it whenever the user picked a speed (the replicate form's 语速 field sends one). Above ~1.3 subtitle alignment drifts and the delivery turns mechanical — shorten the script instead. | `1.0` |
-| `--variant` | Template visual variant id (a key of the template's `variants`; `template-registry --list-templates` prints them as `variantIds`). Written to `meta.templateVariant`, which takes precedence over the fuzzy `--style` match. Pass it whenever the user picked a style — the replicate card shows it as `名字（id）`, e.g. `视觉风格: 紫色科技（tech-purple）`; pass the id in the parentheses. Requires `--template-id`; unknown ids are rejected. | template default |
+| `--variant` | Template visual variant id (a key of the template's `variants`; `template-registry --list-templates` prints them as `variantIds`). Written to `meta.templateVariant`, which takes precedence over the fuzzy `--style` match, and the variant's own `imageStyleGuide` is what the skeleton's image prompts / `negativePrompt` are built from (pass it to `gen_script` rather than setting `meta.templateVariant` afterwards, or the prompts keep the template's default look). Pass it whenever the user picked a style — the replicate card shows it as `名字（id）`, e.g. `视觉风格: 紫色科技（tech-purple）`; pass the id in the parentheses. Requires `--template-id`; unknown ids are rejected. | template default |
 | `--background-preset` | Scene background preset, one of the template's `customPayloadSchema.background.preset` values. Written to every scene's `customPayload.background.preset` (other keys of the template's default background are kept). Pass it whenever the user picked a background, e.g. `背景样式: 纯色（solid）` → `solid`. `image` / `video` also need `customPayload.background.assetRef`. Requires `--template-id`; unsupported values are rejected. | template default |
 | `--font-id` | Font library key (`font.uniq_id`), written to `global.font.fontId`; ab-render resolves and installs the real font at render time. Pass it whenever the user picked a font — the replicate card shows it as `名字（id）`, e.g. `字体: <名字>（<uniqId>）`; pass the id in the parentheses. Never pass a CSS family or the display name: an unknown key falls back to the template font with no error. | template font |
 | `--font-name` | Display name of the picked font → `global.font.fontName` (display only, never used for rendering). Pass it with `--font-id`. | — |
@@ -482,7 +488,7 @@ python3 <SkillDir>/scripts/gen_script.py \
 | `--template-candidates` | Repeatable. Template ids offered to the user, recommended first → `meta.templateSelection.candidates`. With `user-confirmed`, `--template-id` must be one of them. | — |
 | `--allow-digital-human` | Whether to allow digital-human assets. | off |
 | `--allow-ai-video` | Whether to allow AI-generated video assets. | off |
-| `--validate` | Validate-only mode: only validate the input DSL. | — |
+| `--validate` | Validate-only mode: validate the input DSL against the DSL schema **and** the template contract (customPayload field names, enum values, asset references). Warnings are printed but do not fail. | — |
 | `--input` | Input DSL file path (required in validate mode). | — |
 | `-o` / `--output` | Output DSL file path. | stdout |
 | `--stub-image-url` | Test mode: every image AssetRef is written as existing + generated + this URL, no prompt (env: `STUB_IMAGE_URL`). | — |

package/skills/gen-script/scripts/gen_script.py

@@ -302,6 +302,33 @@ def validate_dsl(dsl: dict) -> list:
     return errors_as_strings(validate_structural(dsl))
 
 
+def validate_authored_dsl(dsl: dict) -> tuple[list, list]:
+
+
+
+
+
+
+
+
+
+    from video_dsl.runtime.dsl_validator import (
+        validate_integrity,
+        validate_structural,
+        errors_as_strings,
+    )
+    found, seen = [], set()
+
+    for e in validate_structural(dsl) + validate_integrity(dsl):
+        key = (e.path, e.message)
+        if key not in seen:
+            seen.add(key)
+            found.append(e)
+    errors = [e for e in found if getattr(e, "severity", "error") == "error"]
+    warnings = [e for e in found if getattr(e, "severity", "error") != "error"]
+    return errors_as_strings(errors), errors_as_strings(warnings)
+
+
 def plan_scenes(
     topic: str,
     duration: int,
@@ -609,6 +636,8 @@ def _template_needs_image(template_config: dict | None) -> bool:
 
 
 
+
+
 
 
 
@@ -674,6 +703,7 @@ def _resolve_contract(template_config: dict | None) -> dict:
         "duration_model": caps.get("durationModel"),
         "payload_defaults": payload_defaults,
         "primary_visual_type": primary_visual_type,
+        "background_strategy": "shared" if caps.get("backgroundStrategy") == "shared" else "per-scene",
         "default_layout": caps.get("defaultLayout"),
         "authored_scenes": caps.get("authoredScenes") is True,
 
@@ -1518,6 +1548,7 @@ def build_dsl(
         scene_images: list[str] | None = None,
         item_count: int | None = None,
         source_document_url: str | None = None,
+        variant: str | None = None,
 ) -> dict:
 
 
@@ -1708,16 +1739,33 @@ def build_dsl(
 
 
 
-    image_style_guide = resolve_style_guide(template_config) if _HAS_ENHANCER else {}
+
+
+
+    image_style_guide = (
+        resolve_style_guide(template_config, variant or "") if _HAS_ENHANCER else {}
+    )
     image_compose = _HAS_ENHANCER and is_compose_mode(image_style_guide)
     image_ratio = image_style_guide.get("ratio") or ratio
 
     tail_pad_sec = _nb.template_tail_pad_sec(template_config) if _HAS_BUDGET else None
+
+
+
+
+    shared_background = (
+        contract["background_strategy"] == "shared"
+        and primary_visual_type == "image"
+        and not provided_slots
+    )
+    shared_background_planned = False
     for idx, plan in enumerate(scene_plans):
         scene_id = f"scene-{idx + 1:02d}"
 
         if primary_visual_type == "video":
             visual_asset_id = f"video-{scene_id}"
+        elif shared_background:
+            visual_asset_id = "img-background"
         else:
             visual_asset_id = f"img-{scene_id}"
         narration_asset_id = f"narration-{scene_id}" if needs_narration else None
@@ -1733,7 +1781,11 @@ def build_dsl(
         scene_has_visual = has_visual and _slot_reads_visual(template_config, plan["purpose"])
 
 
-        if scene_has_visual and primary_visual_type == "image":
+        if scene_has_visual and shared_background and shared_background_planned:
+            pass
+        elif scene_has_visual and primary_visual_type == "image":
+            if shared_background:
+                shared_background_planned = True
             slot = provided_slots[idx] if idx < len(provided_slots) else ""
             slot_is_ai, slot_prompt = _parse_ai_slot(slot)
             if slot and not slot_is_ai:
@@ -1760,6 +1812,12 @@ def build_dsl(
             else:
                 if image_compose:
                     image_result = {"prompt": COMPOSE_SUBJECT_PLACEHOLDER}
+                elif shared_background:
+
+
+                    image_result = generate_image_prompt(
+                        "point", topic, style, narration_text="", style_guide=image_style_guide,
+                    )
                 else:
                     image_result = generate_image_prompt(
                         plan["purpose"], topic, style,
@@ -2082,7 +2140,8 @@ Examples:
         "--variant", default=None,
         help="Template visual variant id (a key of the template's `variants`, listed as variantIds by "
              "template-registry --list-templates). Written to meta.templateVariant, which overrides the "
-             "fuzzy --style match. When the user picked a style in the replicate card (shown as "
+             "fuzzy --style match, and the variant's imageStyleGuide drives the skeleton's image "
+             "prompts. When the user picked a style in the replicate card (shown as "
              "`名字（id）`, e.g. `紫色科技（tech-purple）`), pass that id. Requires --template-id.",
     )
     parser.add_argument(
@@ -2189,7 +2248,9 @@ Examples:
             sys.exit(1)
         with open(args.input, "r", encoding="utf-8") as f:
             dsl = json.load(f)
-        errors = validate_dsl(dsl)
+        errors, warnings = validate_authored_dsl(dsl)
+        for w in warnings:
+            print(f"⚠️  {w}", file=sys.stderr)
         if errors:
             print("❌ DSL validation failed:", file=sys.stderr)
             for err in errors:
@@ -2316,6 +2377,7 @@ Examples:
             scene_images=args.scene_images,
             item_count=args.item_count,
             source_document_url=args.source_document_url,
+            variant=args.variant,
         )
     except ValueError as exc:
 

package/skills/gen-script/version.json

@@ -2,6 +2,6 @@
   "skillName": "gen-script",
   "repoName": "agent-skill-media-maker",
   "skillId": "474",
-  "version": "V21",
+  "version": "V22",
   "skillDescription": "Video-script generation skill. Turns a user-supplied topic into a structured Video DSL (JSON) that describes the full video — scene structure, asset requirements, and narrative flow.\n\nUse this skill as soon as the user mentions any of these intents:\n- Write a video script, generate a video script, plan a video, write storyboards\n- Create a short video, plan video content, generate a Video DSL\n- Turn a topic into a video structure / video content plan\n\nEven when the user does not say \"generate the DSL\", use this skill whenever they want to turn a topic into a structured video plan.\n\n⚠️ Stop-and-confirm gate: after this skill returns a DSL, show the full script and wait for the user's explicit confirmation. Never call `prepare_video_assets` in the same turn."
 }

package/skills/render-video/scripts/render_video.py

@@ -779,6 +779,12 @@ def _probe_video_duration(url: str):
     return probe_video_duration(url)
 
 
+def _probe_video_meta(url: str) -> dict:
+
+    from _video_probe import probe_video_meta
+    return probe_video_meta(url)
+
+
 def build_render_plan(dsl: dict, binding: dict) -> dict:
 
     fps = dsl.get("global", {}).get("fps", 30)
@@ -868,11 +874,35 @@ def build_render_plan(dsl: dict, binding: dict) -> dict:
 
 
 
+
+
+
+
+
+
+
+    meta_by_url: dict = {}
     for aid, asset in assets_by_id.items():
-        if asset["type"] == "video" and asset["source"] == "existing" and not asset.get("duration") and asset.get("url"):
-            probed = _probe_video_duration(asset["url"])
-            if probed:
-                asset["duration"] = probed
+        url = asset.get("url")
+        if asset["type"] != "video" or asset["source"] != "existing" or not url:
+            continue
+        need_size = not (asset.get("width") and asset.get("height"))
+        need_duration = not asset.get("duration")
+        if not (need_size or need_duration):
+            continue
+        meta = meta_by_url.get(url)
+        if meta is None and need_size:
+            meta = meta_by_url[url] = _probe_video_meta(url)
+        meta = meta or {}
+        if need_size and meta.get("width") and meta.get("height"):
+            asset["width"], asset["height"] = meta["width"], meta["height"]
+        if need_duration:
+            if meta.get("duration_ms"):
+                asset["duration"] = meta["duration_ms"] / 1000.0
+            else:
+                probed = _probe_video_duration(url)
+                if probed:
+                    asset["duration"] = probed
 
     timeline = []
     current_frame = 0

package/skills/template-registry/scripts/registry_loader.py

@@ -365,6 +365,11 @@ def _allowed_statuses() -> set[str]:
 
 
 
+
+
+
+
+
     allowed = {"stable"}
     if _truthy_env(_BETA_GATE_ENV):
         allowed.add("beta")

package/skills/template-registry/version.json

@@ -2,6 +2,6 @@
   "skillName": "template-registry",
   "repoName": "agent-skill-media-maker",
   "skillId": "475",
-  "version": "V28",
+  "version": "V29",
   "skillDescription": "Video-template registry skill. Stores every video-template definition and lists the available templates (templateId / name / aspect ratio / style tags).\n\nUse this skill as soon as the user mentions any of these intents:\n- View available templates / list every template\n\nNote: DSL→TemplateBinding is no longer a separate exposed step — once prepare_video_assets receives a template_id it builds the binding internally."
 }

package/skills/template-registry/video_dsl/runtime/dsl_validator.py

@@ -1014,6 +1014,10 @@ def _walk_payload(
 
 
 
+
+
+
+
     errors = _check_payload_object(
         data,
         schema,
@@ -1086,6 +1090,10 @@ def _check_payload_object(
                 "the slide renders its built-in placeholder instead of your content.",
             ))
 
+    for key, sub in merged_props.items():
+        if key in data and isinstance(sub, dict):
+            errors += _enum_errors(data[key], sub, path=f"{path}.{key}", scene=scene, key=key)
+
 
 
     for key, sub in props.items():
@@ -1104,6 +1112,41 @@ def _check_payload_object(
     return errors
 
 
+def _enum_errors(
+    value: object, schema: dict, *, path: str, scene: str, key: str
+) -> list[ValidationError]:
+
+
+
+
+
+    enum = schema.get("enum")
+    if isinstance(enum, list) and enum:
+        if value is None or isinstance(value, (dict, list)) or value in enum:
+            return []
+        hint = ""
+        if isinstance(value, str):
+            hit = difflib.get_close_matches(
+                value, [e for e in enum if isinstance(e, str)], n=1, cutoff=0.6
+            )
+            if hit:
+                hint = f" Did you mean '{hit[0]}'?"
+        allowed = ", ".join(repr(e) for e in enum)
+        return [ValidationError(
+            path,
+            f"scene '{scene}': '{key}' is {value!r}, which is not one of [{allowed}] — the "
+            "template ignores an unknown value and falls back to its default, so the look "
+            "you asked for silently never appears." + hint,
+        )]
+    items = schema.get("items")
+    if schema.get("type") == "array" and isinstance(items, dict) and isinstance(value, list):
+        errors: list[ValidationError] = []
+        for i, item in enumerate(value):
+            errors += _enum_errors(item, items, path=f"{path}[{i}]", scene=scene, key=f"{key}[{i}]")
+        return errors
+    return []
+
+
 def _is_blank(value: object) -> bool:
 
 

package/skills/template-registry/video_dsl/runtime/payload_contract.py

@@ -59,13 +59,22 @@ def required_payload_errors(template: dict, props: dict) -> list[str]:
 
 def asset_references(value, path: str = "", *, bound_props: bool = False):
 
+
+
+
+
+
     if isinstance(value, list):
         for index, item in enumerate(value):
             yield from asset_references(item, f"{path}[{index}]", bound_props=bound_props)
     elif isinstance(value, dict):
         for key, child in value.items():
             child_path = f"{path}.{key}" if path else key
-            if key == "assetRef" or (bound_props and key.endswith("AssetId")):
+            if (
+                key == "assetRef"
+                or key.endswith("AssetRef")
+                or (bound_props and key.endswith("AssetId"))
+            ):
                 if isinstance(child, str) and child.strip():
                     yield child_path, child
             else:

package/skills/template-registry/video_dsl/runtime/prompt_enhancer.py

@@ -462,7 +462,9 @@ def compose_dsl_image_prompts(dsl: dict, template: Optional[dict]) -> list[str]:
 
 
     meta = dsl.get("meta") or {}
-    variant_id = str(meta.get("variantId") or "")
+
+
+    variant_id = str(meta.get("templateVariant") or meta.get("variantId") or "")
     if not is_compose_mode(resolve_style_guide(template, variant_id)):
         return []