@remixmate/cli 0.12.8 → 0.12.9

package/dist/manifest.json

@@ -1,7 +1,7 @@
 {
   "schemaVersion": 1,
-  "version": "0.12.8",
-  "generatedAt": "2026-10-20T23:46:18.482Z",
+  "version": "0.12.9",
+  "generatedAt": "2026-10-21T03:02:36.247Z",
   "skills": [
     {
       "id": "doc-parse",
@@ -625,6 +625,17 @@
             ],
             "description": "Camera motion for every scene (scenes[].animationHints.motion), replacing the skeleton's alternating push-in / pull-out. **Must** pass when the user picked one: `运镜效果: 缓慢推近（kenburns-in）` → `kenburns-in`."
           },
+          "ending_payload": {
+            "type": "array",
+            "items": {
+              "type": "string"
+            },
+            "description": "Like payload, but `KEY=VALUE` is written only to the ending scene and locked there (e.g. `recap.cta=关注我，每周更新开源热榜` for github-repo-rank's 片尾引导). On countdown templates (Top-N lists) it also appends the ending scene after #1 — fill that scene's narration (a closing line) and its customPayload.item with the #1 entry. Without it a countdown ends on the #1 card. Requires template_id."
+          },
+          "avatar_id": {
+            "type": "string",
+            "description": "Digital-human avatar id (`dh_xxx`, the replicate card's 数字人主讲 / 数字人讲师 line) for templates declaring capabilities.digitalHuman (image-slide, ppt-to-video, html-slide, …). Adds one gen-digital-human asset referenced by every scene's visuals.avatar.assetRef: one continuous presenter whose lips follow the narration. Placement stays the template default; adjust with payload `digitalHuman.<field>=…`. Copy the id verbatim, never invent one. Costs digital-human credits, so pass it only when the user picked an avatar. Requires template_id."
+          },
           "font_id": {
             "type": "string",
             "description": "Font library key (font.uniq_id) for the on-screen text and subtitles. Written to global.font.fontId; match_template folds it into the binding's typography and ab-render resolves and installs the real font before rendering. **Must** pass when 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. Pass the opaque key, never a CSS family name or the display name — those fall back to the template font with no error. When the user names a font without its id, call gen_script with list_fonts=true first and take the id from that list; do not invent ids. Without it the choice is silently dropped and the video renders in the template's own font. Omit to follow the template's font pairing."
@@ -728,6 +739,8 @@
           "sequence_fit",
           "background_asset",
           "payload",
+          "ending_payload",
+          "avatar_id",
           "motion",
           "font_id",
           "font_name",

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@remixmate/cli",
-  "version": "0.12.8",
+  "version": "0.12.9",
   "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

@@ -486,6 +486,8 @@ python3 <SkillDir>/scripts/gen_script.py \
 | `--sequence-fit` | Image-sequence fill mode, one of `customPayloadSchema.sequence.fit` (`cover` fill and crop / `contain` whole picture over a blurred backdrop), written to `customPayload.sequence.fit`; beats the preset's fit. `auto` = template / preset default. | template / preset default |
 | `--background-asset` | URL of the user's background image or video (`.mp4/.mov/.webm/.m4v` → video). Added as existing asset `asset-background`; every scene's `customPayload.background` gets `preset: image` + `assetRef`, or `preset: video` + `videoAssetRef`. Conflicts with a different `--background-preset`; rejected when the template's background enum lacks that kind. | none |
 | `--payload KEY=VALUE` | Template-specific card choice, repeatable. `KEY` is a dotted path in `customPayloadSchema` (or `capabilities.frameLevelKeys`), e.g. `grade=四年级`, `watermarkText=…`, `preset=course`, `carousel.fit=cover`, `caption.charsPerSec=20`, `templateData.revealMode=progressive`. Value is converted to the declared type (true/false, numbers) and checked against enum / range; undeclared keys are rejected. Written to every scene with sibling keys kept, applied before fit-caption duration estimation. | template default |
+| `--ending-payload KEY=VALUE` | Same as `--payload`, but written only to the ending scene, e.g. `recap.cta=关注我，每周更新开源热榜`. On countdown templates it also appends an ending scene after #1 (it carries `item.rank=1`; fill its narration and `item`). Errors when the script has no ending scene. | none |
+| `--avatar-id` | Digital-human avatar id (`dh_xxx`) for templates declaring `capabilities.digitalHuman`. Adds asset `asset-avatar` (`type: avatar`, `source: gen-digital-human`, `payload.avatarId`) and sets `visuals.avatar.assetRef` on every scene whose slot reads it — one continuous presenter driven by the narration. Placement: template default, or `--payload digitalHuman.<field>=…` (a photo-only presenter is `--payload digitalHuman.posterUrl=<url>` without this flag). | none |
 | `--motion` | Camera motion for every scene (`animationHints.motion`): `static` / `kenburns-in` / `kenburns-out` / `pan-left` / `pan-right` / `float`. | skeleton alternates in / out |
 | `--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`. | — |
@@ -503,7 +505,7 @@ python3 <SkillDir>/scripts/gen_script.py \
 | `--scene-images` | Repeatable. One scene's visual for a **`visual-overlay`** template, mapped **positionally**. Each entry is a URL (that existing asset) or `ai` / `ai:<prompt>` (that scene goes to gen-image, optionally with the user's own prompt) — so an AI scene can sit anywhere in the order. Scenes past the last entry still get AI images. Scene count is raised (never lowered) to fit the entries when `--scenes` is absent. **Required** whenever the user supplies images for such a template — omitting it silently ignores every image they gave. | — |
 | `--caption-lines` | Repeatable. On-screen caption line for `carousel-caption` templates → `customPayload.caption.lines`. Line limits and `**emphasis**` support are declared per template (see its `llmHint`). **Required** for `durationStrategy: fit-caption` templates — write the lines yourself if the user did not supply them. Optional only for `fit-images` templates. | — |
 
-Values written by `--background-preset`, `--background-asset`, `--payload` and `--motion` are listed in `meta.lockedScenePaths` (scene-relative paths such as `customPayload.carousel.fit`, `animationHints.motion`). When the agent later saves its own `dsl_json` in merge mode, ab-agent restores those paths from this skeleton after merging, so a model-written `carousel` / `templateData` / `background` object cannot wipe a card choice. Do not hand-write these keys in `dsl_json`; pass the flag.
+Values written by `--background-preset`, `--background-asset`, `--payload`, `--ending-payload`, `--avatar-id` and `--motion` are listed in `meta.lockedScenePaths` (scene-relative paths such as `customPayload.carousel.fit`, `animationHints.motion`). When the agent later saves its own `dsl_json` in merge mode, ab-agent restores those paths from this skeleton after merging, so a model-written `carousel` / `templateData` / `background` object cannot wipe a card choice. Do not hand-write these keys in `dsl_json`; pass the flag. Templates that author their own scenes (`meta.authoredScenes`, e.g. screen-walkthrough) are not merged: there ab-agent writes each locked value that the skeleton holds on every scene (with one value) onto every submitted scene, and adds the skeleton assets those values reference (`asset-background`, `asset-avatar`) when the submitted `assets` lacks them.
 
 ## DSL generation principles
 

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

@@ -1430,7 +1430,12 @@ def _build_custom_payload(template_config: dict | None) -> dict:
     return {}
 
 
-def _plan_countdown_scenes(spec: dict, count: int, duration: float, headline: str) -> list:
+def _plan_countdown_scenes(
+    spec: dict, count: int, duration: float, headline: str, ending: bool = False
+) -> list:
+
+
+
 
 
 
@@ -1440,11 +1445,14 @@ def _plan_countdown_scenes(spec: dict, count: int, duration: float, headline: st
 
 
     head_sec = countdown_opening_sec(spec, count, duration)
-    beat_sec = (duration - head_sec) / count
-    return [{"purpose": "opening", "duration": head_sec, "label": headline}] + [
+    beat_sec = (duration - head_sec) / (count + (1 if ending else 0))
+    plans = [{"purpose": "opening", "duration": head_sec, "label": headline}] + [
         {"purpose": "point", "duration": beat_sec, "label": f"#{index}", "countdownIndex": index}
         for index in range(count, 0, -1)
     ]
+    if ending:
+        plans.append({"purpose": "ending", "duration": beat_sec, "label": "#1", "countdownIndex": 1})
+    return plans
 
 
 def _warn_on_count_mismatch(spec: dict, headline: str, subheadline: str, count: int) -> None:
@@ -1797,6 +1805,102 @@ def _write_payload(custom_payload: dict, writes: PayloadWrites) -> None:
         node[parts[-1]] = value
 
 
+def _template_has_ending_slot(template_config: dict | None) -> bool:
+    return isinstance(((template_config or {}).get("slotMapping") or {}).get("ending"), dict)
+
+
+def _apply_ending_payload(
+    dsl: dict,
+    template_config: dict | None,
+    template_id: str | None,
+    payload_specs: list[str] | None,
+) -> None:
+
+
+
+
+
+
+
+    if not payload_specs:
+        return
+    if not template_id or template_config is None:
+        raise ValueError("--ending-payload needs --template-id: the allowed keys come from the template.")
+    try:
+        writes = _parse_payload_specs(template_config, payload_specs)
+    except ValueError as exc:
+        raise ValueError(str(exc).replace("--payload ", "--ending-payload ", 1)) from None
+    endings = [s for s in dsl.get("scenes", []) if s.get("purpose") == "ending"]
+    if not endings:
+        raise ValueError(
+            "--ending-payload: this script has no ending scene to write to "
+            f"(scene purposes: {', '.join(str(s.get('purpose')) for s in dsl.get('scenes', []))})."
+        )
+    for scene in endings:
+        _write_payload(scene.setdefault("customPayload", {}), writes)
+    for parts, _ in writes:
+        _lock_scene_path(dsl, "customPayload." + ".".join(parts))
+
+
+AVATAR_ASSET_ID = "asset-avatar"
+
+
+def _slot_extracts(template_config: dict | None, purpose: str, source: str) -> bool:
+
+    mapping = (template_config or {}).get("slotMapping")
+    if not isinstance(mapping, dict):
+        return False
+    slot = mapping.get(purpose) or mapping.get("default") or {}
+    seen: set = set()
+    while isinstance(slot, dict) and "$ref" in slot and slot["$ref"] not in seen:
+        seen.add(slot["$ref"])
+        slot = mapping.get(slot["$ref"]) or {}
+    extractors = slot.get("propExtractors") if isinstance(slot, dict) else None
+    return isinstance(extractors, dict) and any(
+        isinstance(ex, dict) and ex.get("from") == source for ex in extractors.values()
+    )
+
+
+def _apply_avatar(
+    dsl: dict,
+    template_config: dict | None,
+    template_id: str | None,
+    avatar_id: str | None,
+) -> None:
+
+
+
+
+
+
+    if not avatar_id:
+        return
+    if not template_id or template_config is None:
+        raise ValueError("--avatar-id needs --template-id: the template decides where the presenter shows.")
+    support = ((template_config.get("capabilities") or {}).get("digitalHuman") or {}).get("support")
+    if support not in ("optional", "required"):
+        raise ValueError(f"--avatar-id: template {template_id!r} does not declare capabilities.digitalHuman")
+    scenes = [
+        s for s in dsl.get("scenes", [])
+        if _slot_extracts(template_config, str(s.get("purpose")), "visuals.avatar.assetRef")
+    ]
+    if not scenes:
+        raise ValueError(f"--avatar-id: no scene of template {template_id!r} reads visuals.avatar.assetRef")
+    if not any((s.get("audio") or {}).get("narration") for s in scenes):
+        raise ValueError("--avatar-id: the presenter is driven by the narration, and this script has none")
+    for scene in scenes:
+        visuals = scene.get("visuals") if isinstance(scene.get("visuals"), dict) else {}
+        scene["visuals"] = {**visuals, "avatar": {"assetRef": AVATAR_ASSET_ID}}
+    dsl.setdefault("assets", []).append({
+        "assetId": AVATAR_ASSET_ID,
+        "type": "avatar",
+        "source": "gen-digital-human",
+        "status": "planned",
+        "payload": {"avatarId": avatar_id},
+    })
+    _lock_scene_path(dsl, "visuals.avatar.assetRef")
+
+
 def build_dsl(
         topic: str,
         duration: int,
@@ -1824,6 +1928,7 @@ def build_dsl(
         sequence_preset: str | None = None,
         sequence_fit: str | None = None,
         payload_writes: PayloadWrites | None = None,
+        countdown_ending: bool = False,
 ) -> dict:
 
 
@@ -1862,6 +1967,10 @@ def build_dsl(
         )
     beat_count = resolve_countdown_count(countdown, item_count, scene_count) if countdown else None
 
+    countdown_ending = bool(countdown_ending and countdown)
+    if countdown_ending and not _template_has_ending_slot(template_config):
+        raise ValueError(f"--ending-payload: template {template_id!r} has no ending slot")
+
     primary_visual_type = contract["primary_visual_type"]
     payload_style = contract["payload_style"]
 
@@ -1980,7 +2089,9 @@ def build_dsl(
     provided_slots = [s.strip() for s in (scene_images or []) if s and s.strip()]
 
     if beat_count is not None:
-        scene_plans = _plan_countdown_scenes(countdown, beat_count, duration, effective_headline)
+        scene_plans = _plan_countdown_scenes(
+            countdown, beat_count, duration, effective_headline, ending=countdown_ending
+        )
     else:
         scene_plans = _plan_contract_scenes(contract, topic, duration, scene_count, output_language)
 
@@ -2465,6 +2576,8 @@ Examples:
              "pan-left / pan-right / float. Replaces the skeleton's alternating default.",
     )
     parser.add_argument("--allow-digital-human", action="store_true", help="Allow digital-human assets")
+    parser.add_argument("--avatar-id", default=None, help="Digital-human avatar id (dh_xxx) for templates declaring capabilities.digitalHuman: adds one gen-digital-human asset referenced by every scene's visuals.avatar.assetRef (one continuous presenter driven by the narration). Placement stays the template default unless set with --payload digitalHuman.<field>=...")
+    parser.add_argument("--ending-payload", action="append", default=None, metavar="KEY=VALUE", help="Like --payload, but written only to the ending scene (e.g. recap.cta=关注我，每周更新开源热榜). On countdown templates it also appends the ending scene after #1. Repeatable.")
     parser.add_argument("--allow-ai-video", action="store_true", help="Allow AI-generated video assets")
     parser.add_argument("--template-id", default=None, help="Template id. The template owns outputLanguage and may declare a defaultVoiceId.")
     parser.add_argument(
@@ -2705,6 +2818,7 @@ Examples:
             sequence_preset=args.sequence_preset,
             sequence_fit=args.sequence_fit,
             payload_writes=payload_writes,
+            countdown_ending=bool(args.ending_payload),
         )
     except ValueError as exc:
 
@@ -2723,6 +2837,8 @@ Examples:
             args.background_asset,
         )
         _apply_scene_payload(dsl, look_config, args.template_id, args.payload, args.motion)
+        _apply_ending_payload(dsl, look_config, args.template_id, args.ending_payload)
+        _apply_avatar(dsl, look_config, args.template_id, args.avatar_id)
     except ValueError as exc:
         parser.error(str(exc))
 

package/skills/gen-script/skill.json

@@ -87,6 +87,17 @@
         ],
         "description": "Camera motion for every scene (scenes[].animationHints.motion), replacing the skeleton's alternating push-in / pull-out. **Must** pass when the user picked one: `运镜效果: 缓慢推近（kenburns-in）` → `kenburns-in`."
       },
+      "ending_payload": {
+        "type": "array",
+        "items": {
+          "type": "string"
+        },
+        "description": "Like payload, but `KEY=VALUE` is written only to the ending scene and locked there (e.g. `recap.cta=关注我，每周更新开源热榜` for github-repo-rank's 片尾引导). On countdown templates (Top-N lists) it also appends the ending scene after #1 — fill that scene's narration (a closing line) and its customPayload.item with the #1 entry. Without it a countdown ends on the #1 card. Requires template_id."
+      },
+      "avatar_id": {
+        "type": "string",
+        "description": "Digital-human avatar id (`dh_xxx`, the replicate card's 数字人主讲 / 数字人讲师 line) for templates declaring capabilities.digitalHuman (image-slide, ppt-to-video, html-slide, …). Adds one gen-digital-human asset referenced by every scene's visuals.avatar.assetRef: one continuous presenter whose lips follow the narration. Placement stays the template default; adjust with payload `digitalHuman.<field>=…`. Copy the id verbatim, never invent one. Costs digital-human credits, so pass it only when the user picked an avatar. Requires template_id."
+      },
       "font_id": {
         "type": "string",
         "description": "Font library key (font.uniq_id) for the on-screen text and subtitles. Written to global.font.fontId; match_template folds it into the binding's typography and ab-render resolves and installs the real font before rendering. **Must** pass when 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. Pass the opaque key, never a CSS family name or the display name — those fall back to the template font with no error. When the user names a font without its id, call gen_script with list_fonts=true first and take the id from that list; do not invent ids. Without it the choice is silently dropped and the video renders in the template's own font. Omit to follow the template's font pairing."
@@ -190,6 +201,8 @@
       "sequence_fit",
       "background_asset",
       "payload",
+      "ending_payload",
+      "avatar_id",
       "motion",
       "font_id",
       "font_name",