@remixmate/cli 0.9.54 → 0.9.56

package/dist/handlers/gen-digital-human.d.ts

@@ -1,12 +1,17 @@
 /**
- * gen-digital-human handler — talking-head video, three modes:
- *   --list-avatars true                  → POST /digital-human/avatar/page (list)
- *   --check-status --generation-id <id>  → POST /digital-human/video/status (one-shot)
- *   default (--avatar-id <id> ...)       → generate: resolve avatar → submit → poll
+ * gen-digital-human handler — talking-head video + custom avatars, five modes:
+ *   --list-avatars true                          → POST /digital-human/avatar/page (list)
+ *   --clone --clone-url <url> --name <n>         → POST /digital-human/avatar/clone, then poll clone-status
+ *   --clone-status --avatar-id <id>              → POST /digital-human/avatar/clone-status (one-shot)
+ *   --check-status --generation-id <id>          → POST /digital-human/video/status (one-shot)
+ *   default (--avatar-id <id> ...)               → generate: resolve avatar → submit → poll
  *
  * Ported from `scripts/gen_digital_human_video.py` (python → http) so it shares
  * auth / base-URL / business-code handling with the other generators via
- * http.ts. Endpoints and payloads match the backend DigitalHumanVideoGenerateDTO.
+ * http.ts. Endpoints and payloads match the backend DigitalHuman* DTOs.
+ *
+ * Avatar ids are the backend's opaque uids (`dh_...`); legacy numeric ids are
+ * still accepted as strings.
  */
 import type { HandlerContext, HandlerInput } from './index.js';
 export declare function genDigitalHuman(input: HandlerInput, ctxIn: HandlerContext): Promise<void>;

package/dist/handlers/gen-digital-human.js

@@ -1,17 +1,28 @@
 /**
- * gen-digital-human handler — talking-head video, three modes:
- *   --list-avatars true                  → POST /digital-human/avatar/page (list)
- *   --check-status --generation-id <id>  → POST /digital-human/video/status (one-shot)
- *   default (--avatar-id <id> ...)       → generate: resolve avatar → submit → poll
+ * gen-digital-human handler — talking-head video + custom avatars, five modes:
+ *   --list-avatars true                          → POST /digital-human/avatar/page (list)
+ *   --clone --clone-url <url> --name <n>         → POST /digital-human/avatar/clone, then poll clone-status
+ *   --clone-status --avatar-id <id>              → POST /digital-human/avatar/clone-status (one-shot)
+ *   --check-status --generation-id <id>          → POST /digital-human/video/status (one-shot)
+ *   default (--avatar-id <id> ...)               → generate: resolve avatar → submit → poll
  *
  * Ported from `scripts/gen_digital_human_video.py` (python → http) so it shares
  * auth / base-URL / business-code handling with the other generators via
- * http.ts. Endpoints and payloads match the backend DigitalHumanVideoGenerateDTO.
+ * http.ts. Endpoints and payloads match the backend DigitalHuman* DTOs.
+ *
+ * Avatar ids are the backend's opaque uids (`dh_...`); legacy numeric ids are
+ * still accepted as strings.
  */
 import { mmPost, pollUntil, resolveHttpContext, SkillError } from '../http.js';
 import { getCapabilities } from '../capabilities.js';
 import { emitProgress } from '../progress.js';
 import { isTrue, toNumber } from './shared.js';
+const CLONE_TYPES = ['video', 'image'];
+function toId(value) {
+    if (value === undefined || value === null)
+        return '';
+    return String(value).trim();
+}
 async function listAvatars(ctx, input) {
     const payload = { current: 1, pageSize: 100 };
     if (input.source)
@@ -32,10 +43,10 @@ async function listAvatars(ctx, input) {
         process.stdout.write('⚠️  No avatars available\n');
         return;
     }
-    process.stdout.write(`\n${'ID'.padEnd(8)} ${'Name'.padEnd(16)} ${'Source'.padEnd(10)} ${'Gender'.padEnd(8)} Ratio\n`);
-    process.stdout.write(`${'-'.repeat(8)} ${'-'.repeat(16)} ${'-'.repeat(10)} ${'-'.repeat(8)} ${'-'.repeat(10)}\n`);
+    process.stdout.write(`\n${'ID'.padEnd(24)} ${'Name'.padEnd(16)} ${'Source'.padEnd(14)} ${'Gender'.padEnd(8)} ${'Ratio'.padEnd(6)} Clone\n`);
+    process.stdout.write(`${'-'.repeat(24)} ${'-'.repeat(16)} ${'-'.repeat(14)} ${'-'.repeat(8)} ${'-'.repeat(6)} ${'-'.repeat(8)}\n`);
     for (const a of avatars) {
-        process.stdout.write(`${String(a.id ?? '').padEnd(8)} ${String(a.name ?? '').padEnd(16)} ${String(a.source ?? '').padEnd(10)} ${String(a.gender ?? '').padEnd(8)} ${a.aspectRatio ?? ''}\n`);
+        process.stdout.write(`${toId(a.id).padEnd(24)} ${String(a.name ?? '').padEnd(16)} ${String(a.source ?? '').padEnd(14)} ${String(a.gender ?? '').padEnd(8)} ${String(a.aspectRatio ?? '').padEnd(6)} ${a.cloneStatus || '-'}\n`);
     }
     process.stdout.write(`\n${avatars.length} avatar(s) available\n`);
     process.stdout.write('💡 Use --avatar-id <ID> to pick an avatar for generation\n');
@@ -44,7 +55,7 @@ async function listAvatars(ctx, input) {
 async function fetchAvatarInfo(ctx, avatarId) {
     for (const extra of [{}, { mine: true }]) {
         const data = await mmPost(ctx, '/digital-human/avatar/page', { current: 1, pageSize: 100, ...extra }, { timeoutMs: 30_000 });
-        const found = (data.list ?? []).find((a) => String(a.id) === String(avatarId));
+        const found = (data.list ?? []).find((a) => toId(a.id) === avatarId);
         if (found)
             return found;
     }
@@ -57,6 +68,81 @@ function resolveSource(explicit, avatar, validSources) {
     }
     return src;
 }
+// ── Avatar cloning (HiFly) ─────────────────────────────────────────────────
+async function cloneAvatar(ctx, input) {
+    const fileUrl = toId(input.clone_url);
+    if (!/^https?:\/\//i.test(fileUrl)) {
+        throw new SkillError('❌ --clone needs --clone-url <public http(s) URL of a video or portrait image> (local files are not uploaded)');
+    }
+    const name = toId(input.name);
+    if (!name)
+        throw new SkillError('❌ --clone needs --name <avatar name>');
+    const sourceType = toId(input.clone_type) || 'video';
+    if (!CLONE_TYPES.includes(sourceType)) {
+        throw new SkillError(`❌ --clone-type must be one of ${CLONE_TYPES.join(' / ')}, received: ${sourceType}`);
+    }
+    const payload = { name, sourceType, fileUrl };
+    if (input.thumbnail_url)
+        payload.thumbnailUrl = toId(input.thumbnail_url);
+    if (input.gender)
+        payload.gender = toId(input.gender);
+    if (input.aspect_ratio)
+        payload.aspectRatio = toId(input.aspect_ratio);
+    emitProgress({ phase: 'gen-digital-human:clone-request', sourceType });
+    const data = await mmPost(ctx, '/digital-human/avatar/clone', payload, { timeoutMs: 60_000 });
+    const avatarId = toId(data.id);
+    if (!avatarId) {
+        throw new SkillError(`❌ could not parse the new avatar id; API returned: ${JSON.stringify(data)}`);
+    }
+    emitProgress({ phase: 'gen-digital-human:clone-async', avatarId });
+    const json = isTrue(input.json_output);
+    if (input.wait !== undefined && !isTrue(input.wait)) {
+        if (json) {
+            process.stdout.write(JSON.stringify({ avatarId, cloneStatus: data.cloneStatus ?? 'cloning' }) + '\n');
+            return;
+        }
+        process.stdout.write(`\n🧬 Clone submitted, avatar id: ${avatarId}\n`);
+        process.stdout.write(`💡 Check progress with --clone-status --avatar-id ${avatarId}\n`);
+        return;
+    }
+    let final;
+    try {
+        final = await pollUntil(async () => {
+            const s = await mmPost(ctx, '/digital-human/avatar/clone-status', { id: avatarId }, { timeoutMs: 30_000 });
+            const status = s.cloneStatus ?? 'unknown';
+            emitProgress({ phase: 'gen-digital-human:clone-poll', avatarId, status });
+            if (status === 'done')
+                return s;
+            if (status === 'failed')
+                throw new SkillError(`❌ avatar clone failed: ${s.cloneError || 'unknown error'}`);
+            return null;
+        }, { intervalMs: 10_000, timeoutMs: 20 * 60_000 });
+    }
+    catch (err) {
+        if (err instanceof SkillError && /polling timed out/.test(err.message)) {
+            throw new SkillError(`❌ avatar ${avatarId} is still cloning; rerun --clone-status --avatar-id ${avatarId} later`);
+        }
+        throw err;
+    }
+    if (json) {
+        process.stdout.write(JSON.stringify({ avatarId, cloneStatus: final.cloneStatus }) + '\n');
+        return;
+    }
+    process.stdout.write(`\n✅ Avatar cloned, id: ${avatarId}\n`);
+    process.stdout.write(`💡 Generate with --avatar-id ${avatarId} --text "..."\n`);
+}
+async function checkCloneStatus(ctx, avatarId, json) {
+    const s = await mmPost(ctx, '/digital-human/avatar/clone-status', { id: avatarId }, { timeoutMs: 30_000 });
+    const status = s.cloneStatus || 'unknown';
+    if (json) {
+        process.stdout.write(JSON.stringify({ avatarId, cloneStatus: status, error: s.cloneError || null }) + '\n');
+        return;
+    }
+    process.stdout.write(`\nclone status: ${status}\n`);
+    if (status === 'failed')
+        process.stdout.write(`   failure reason: ${s.cloneError || 'unknown error'}\n`);
+}
+// ── Video generation ───────────────────────────────────────────────────────
 async function checkStatus(ctx, generationId, json) {
     const data = await mmPost(ctx, '/digital-human/video/status', { generationId }, { timeoutMs: 30_000 });
     const status = data.status ?? 'unknown';
@@ -93,21 +179,33 @@ export async function genDigitalHuman(input, ctxIn) {
         privateToken: input.priv_token,
         preflightToken: ctxIn.auth?.token,
     });
+    const json = isTrue(input.json_output);
     // ── Mode 1: list avatars ────────────────────────────────────────────────
     if (isTrue(input.list_avatars)) {
         return listAvatars(ctx, input);
     }
-    // ── Mode 3: check status ────────────────────────────────────────────────
+    // ── Mode 2: clone a custom avatar (HiFly) ───────────────────────────────
+    if (isTrue(input.clone)) {
+        return cloneAvatar(ctx, input);
+    }
+    // ── Mode 3: clone status ────────────────────────────────────────────────
+    if (isTrue(input.clone_status)) {
+        const avatarId = toId(input.avatar_id);
+        if (!avatarId)
+            throw new SkillError('❌ clone-status mode requires --avatar-id');
+        return checkCloneStatus(ctx, avatarId, json);
+    }
+    // ── Mode 4: check generation status ─────────────────────────────────────
     if (isTrue(input.check_status)) {
         const generationId = toNumber(input.generation_id, 'generation_id');
         if (generationId === undefined) {
             throw new SkillError('❌ check-status mode requires --generation-id');
         }
-        return checkStatus(ctx, generationId, isTrue(input.json_output));
+        return checkStatus(ctx, generationId, json);
     }
-    // ── Mode 2: generate (default) ──────────────────────────────────────────
-    const avatarId = toNumber(input.avatar_id, 'avatar_id');
-    if (avatarId === undefined) {
+    // ── Mode 5: generate (default) ──────────────────────────────────────────
+    const avatarId = toId(input.avatar_id);
+    if (!avatarId) {
         throw new SkillError('❌ provide --avatar-id to pick the avatar (use --list-avatars to see available avatars)');
     }
     const audioUrl = (input.audio_url ?? '').trim();
@@ -120,6 +218,12 @@ export async function genDigitalHuman(input, ctxIn) {
     const dh = (await getCapabilities(input.api_base_url)).digitalHuman;
     emitProgress({ phase: 'gen-digital-human:resolve-avatar', avatarId });
     const avatar = await fetchAvatarInfo(ctx, avatarId);
+    if (avatar.cloneStatus === 'cloning') {
+        throw new SkillError(`❌ avatar ${avatarId} is still cloning; check with --clone-status --avatar-id ${avatarId}`);
+    }
+    if (avatar.cloneStatus === 'failed') {
+        throw new SkillError(`❌ avatar ${avatarId} failed to clone (${avatar.cloneError || 'unknown error'}); clone it again`);
+    }
     const source = resolveSource(input.source, avatar, dh.sources);
     const payload = { avatarId, source };
     if (audioUrl) {
@@ -149,7 +253,7 @@ export async function genDigitalHuman(input, ctxIn) {
     if (!videoUrl) {
         throw new SkillError(`❌ video generated but no URL was returned; rerun --check-status --generation-id ${generationId} later`);
     }
-    if (isTrue(input.json_output)) {
+    if (json) {
         process.stdout.write(JSON.stringify({ url: videoUrl, generationId }) + '\n');
         return;
     }

package/dist/manifest.json

@@ -1,7 +1,7 @@
 {
   "schemaVersion": 1,
-  "version": "0.9.54",
-  "generatedAt": "2026-09-28T11:58:29.708Z",
+  "version": "0.9.56",
+  "generatedAt": "2026-09-28T23:55:13.296Z",
   "skills": [
     {
       "id": "doc-parse",
@@ -174,7 +174,8 @@
         "Digital human, talking-head video, AI presenter, virtual host",
         "Make a talking-head video out of this script / this audio file",
         "List or browse digital-human avatars",
-        "Use Jimeng or HiFly to drive an avatar"
+        "Use Jimeng or HiFly to drive an avatar",
+        "Clone / create a custom avatar from a video or photo (HiFly)"
       ],
       "entry": {
         "type": "http",
@@ -199,7 +200,7 @@
           },
           "name": {
             "type": "string",
-            "description": "With list_avatars=true: fuzzy-filter avatars by name"
+            "description": "With list_avatars=true: fuzzy-filter avatars by name. With clone=true: the new avatar's name (required)."
           },
           "source": {
             "type": "string",
@@ -207,7 +208,7 @@
               "jimeng",
               "hifly"
             ],
-            "description": "Provider: jimeng is image-driven, hifly is video-driven. Usually inferred from the avatar; pass it explicitly when the avatar declares no source."
+            "description": "Provider filter / override. hifly avatars (including cloned custom avatars) are video-driven; jimeng is image-driven. Usually inferred from the avatar; pass it explicitly only when the avatar declares no source."
           },
           "gender": {
             "type": "string",
@@ -218,8 +219,36 @@
             "description": "Filter by gender"
           },
           "avatar_id": {
-            "type": "number",
-            "description": "Avatar id"
+            "type": "string",
+            "description": "Avatar id as returned by list_avatars (opaque string like dh_xxx). Also used with clone_status=true."
+          },
+          "clone": {
+            "type": "boolean",
+            "description": "Clone mode: create a custom HiFly avatar from a video (free) or a portrait image (costs HiFly credits). Requires clone_url and name. Waits until cloning finishes unless wait=false."
+          },
+          "clone_url": {
+            "type": "string",
+            "description": "With clone=true: public http(s) URL of the source. Video: mp4/mov (h264), 5s-30min, 360p-4K, ≤500MB, one clearly visible speaking person. Image: a front-facing portrait."
+          },
+          "clone_type": {
+            "type": "string",
+            "enum": [
+              "video",
+              "image"
+            ],
+            "description": "With clone=true: source type. Defaults to video."
+          },
+          "thumbnail_url": {
+            "type": "string",
+            "description": "With clone=true: optional cover image URL shown in avatar lists (image clones default to the image itself)."
+          },
+          "wait": {
+            "type": "boolean",
+            "description": "With clone=true: set false to return right after submission (then poll with clone_status=true). Defaults to waiting up to ~20 minutes."
+          },
+          "clone_status": {
+            "type": "boolean",
+            "description": "Clone-status mode: report cloning / done / failed for avatar_id."
           },
           "text": {
             "type": "string",
@@ -255,7 +284,7 @@
           },
           "json_output": {
             "type": "boolean",
-            "description": "Emit a JSON result (generate: { url, generationId }; list: { avatars }; check-status: { status, url }) instead of human-readable output"
+            "description": "Emit a JSON result (generate: { url, generationId }; list: { avatars }; check-status: { status, url }; clone / clone-status: { avatarId, cloneStatus }) instead of human-readable output"
           }
         },
         "required": []
@@ -275,6 +304,12 @@
         "hidden": [
           "json_output",
           "list_avatars",
+          "clone",
+          "clone_url",
+          "clone_type",
+          "thumbnail_url",
+          "wait",
+          "clone_status",
           "mine",
           "name",
           "gender",

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@remixmate/cli",
-  "version": "0.9.54",
+  "version": "0.9.56",
   "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-digital-human/SKILL.md

@@ -8,6 +8,7 @@ description: |
   - "Make a talking-head video out of this script / this audio file"
   - List or browse digital-human avatars
   - Use Jimeng or HiFly to drive an avatar
+  - Clone / create a custom avatar from a video or photo (HiFly)
 
   Even without an explicit "use AI", any request that turns text or audio into a synthesized presenter video should route here.
 triggers:
@@ -15,6 +16,7 @@ triggers:
   - "Make a talking-head video out of this script / this audio file"
   - List or browse digital-human avatars
   - Use Jimeng or HiFly to drive an avatar
+  - Clone / create a custom avatar from a video or photo (HiFly)
 ---
 
 # Digital-Human Skill
@@ -31,6 +33,10 @@ Wraps ab-api's digital-human endpoints (the same ones the web studio uses), auth
 | TTS lip-sync | avatar + text + voice → talking-head video | `POST /digital-human/video/generate` |
 | Audio-driven | avatar + existing audio URL → talking-head video | `POST /digital-human/video/generate` |
 | Check status | Poll the progress of a generation job | `POST /digital-human/video/status` |
+| Clone avatar | Create a custom HiFly avatar from a video (free) or portrait image (costs HiFly credits) | `POST /digital-human/avatar/clone` |
+| Clone status | Report `cloning` / `done` / `failed` for a cloned avatar | `POST /digital-human/avatar/clone-status` |
+
+Avatar ids are opaque strings (e.g. `dh_xxx`) — always copy them from `--list-avatars`, never invent or renumber them.
 
 **Cross-skill dependency**: voice ids are shared with `gen-voice` (same Minimax catalog). Use `gen-voice --list-voices` to see available voices.
 
@@ -82,7 +88,7 @@ remixmate gen-digital-human --list-avatars --name "alice"
 
 ```bash
 remixmate gen-digital-human \
-  --avatar-id 7 \
+  --avatar-id dh_xxx \
   --text "Hi everyone, welcome to the live stream — today I'll introduce a new product." \
   --voice-id "male-qn-qingse"
 ```
@@ -91,7 +97,7 @@ remixmate gen-digital-human \
 
 ```bash
 remixmate gen-digital-human \
-  --avatar-id 7 \
+  --avatar-id dh_xxx \
   --text "Hi class, here's our brand-new course." \
   --voice-id "female-shaonv" \
   --aspect-ratio "9:16"
@@ -101,7 +107,7 @@ remixmate gen-digital-human \
 
 ```bash
 remixmate gen-digital-human \
-  --avatar-id 39 \
+  --avatar-id dh_yyy \
   --source hifly \
   --audio-url "https://example.com/voice.mp3" \
   --aspect-ratio "16:9"
@@ -113,6 +119,24 @@ remixmate gen-digital-human \
 remixmate gen-digital-human --check-status --generation-id 123
 ```
 
+### 4. Clone a custom avatar (HiFly)
+
+The source must be a public http(s) URL (local files are not uploaded). Video: mp4/mov (h264), 5 s – 30 min, 360p – 4K, ≤ 500 MB, one person speaking to camera. Cloning takes minutes; the handler waits (up to ~20 min) and prints the new avatar id. The avatar also shows up under 素材库 → 数字人 in the web app.
+
+```bash
+# Video clone (free)
+remixmate gen-digital-human --clone --name "My presenter" --clone-url "https://example.com/me.mp4"
+
+# Portrait image clone (costs HiFly credits)
+remixmate gen-digital-human --clone --clone-type image --name "My presenter" --clone-url "https://example.com/me.jpg"
+
+# Submit without waiting, then poll
+remixmate gen-digital-human --clone --wait false --name "My presenter" --clone-url "https://example.com/me.mp4"
+remixmate gen-digital-human --clone-status --avatar-id dh_xxx
+```
+
+A cloned avatar is a normal `hifly` avatar: find it with `--list-avatars --mine`, then generate with `--avatar-id <id> --text "..."` (TTS is synthesized with the gen-voice catalog, then drives the avatar) or `--audio-url`. Audio must be 5 s – 30 min.
+
 ## Common CLI flags
 
 ### List avatars
@@ -147,6 +171,19 @@ remixmate gen-digital-human --check-status --generation-id 123
 | `--check-status` | Enter status-check mode | — |
 | `--generation-id` | Job id (required) | — |
 
+### Clone avatar
+
+| Flag | Description | Default |
+|------|-------------|---------|
+| `--clone` | Enter clone mode | — |
+| `--name` | New avatar name (required) | — |
+| `--clone-url` | Public URL of the source video / image (required) | — |
+| `--clone-type` | `video` / `image` | `video` |
+| `--thumbnail-url` | Cover image shown in lists | image itself / none |
+| `--gender` | `male` / `female` | none |
+| `--wait` | `false` to return right after submission | wait |
+| `--clone-status` | Report clone status of `--avatar-id` | — |
+
 ## Typical flows
 
 ### Flow A — TTS
@@ -196,5 +233,7 @@ generation cost, and the balance is the only warning before a run fails with
 - **Empty TTS text**: provide non-empty `--text` or use `--audio-url`.
 - **Invalid voice id**: re-check via `gen-voice --list-voices`.
 - **Avatar id not found**: re-check via `--list-avatars` (or `--list-avatars --mine`).
+- **Avatar still cloning / clone failed**: poll `--clone-status --avatar-id <id>`; on failure, re-clone with a clearer source (single speaker facing the camera, h264 mp4).
+- **HiFly audio too short**: HiFly needs ≥ 5 s of speech; lengthen the text or audio.
 - **Cannot infer source**: when the avatar lacks a `source` and `--source` is not set, pass `--source jimeng` or `--source hifly` explicitly.
 - **Network**: verify connectivity and `MM_API_BASE_URL`.

package/skills/gen-digital-human/skill.json

@@ -29,7 +29,7 @@
       },
       "name": {
         "type": "string",
-        "description": "With list_avatars=true: fuzzy-filter avatars by name"
+        "description": "With list_avatars=true: fuzzy-filter avatars by name. With clone=true: the new avatar's name (required)."
       },
       "source": {
         "type": "string",
@@ -37,7 +37,7 @@
           "jimeng",
           "hifly"
         ],
-        "description": "Provider: jimeng is image-driven, hifly is video-driven. Usually inferred from the avatar; pass it explicitly when the avatar declares no source."
+        "description": "Provider filter / override. hifly avatars (including cloned custom avatars) are video-driven; jimeng is image-driven. Usually inferred from the avatar; pass it explicitly only when the avatar declares no source."
       },
       "gender": {
         "type": "string",
@@ -48,8 +48,36 @@
         "description": "Filter by gender"
       },
       "avatar_id": {
-        "type": "number",
-        "description": "Avatar id"
+        "type": "string",
+        "description": "Avatar id as returned by list_avatars (opaque string like dh_xxx). Also used with clone_status=true."
+      },
+      "clone": {
+        "type": "boolean",
+        "description": "Clone mode: create a custom HiFly avatar from a video (free) or a portrait image (costs HiFly credits). Requires clone_url and name. Waits until cloning finishes unless wait=false."
+      },
+      "clone_url": {
+        "type": "string",
+        "description": "With clone=true: public http(s) URL of the source. Video: mp4/mov (h264), 5s-30min, 360p-4K, ≤500MB, one clearly visible speaking person. Image: a front-facing portrait."
+      },
+      "clone_type": {
+        "type": "string",
+        "enum": [
+          "video",
+          "image"
+        ],
+        "description": "With clone=true: source type. Defaults to video."
+      },
+      "thumbnail_url": {
+        "type": "string",
+        "description": "With clone=true: optional cover image URL shown in avatar lists (image clones default to the image itself)."
+      },
+      "wait": {
+        "type": "boolean",
+        "description": "With clone=true: set false to return right after submission (then poll with clone_status=true). Defaults to waiting up to ~20 minutes."
+      },
+      "clone_status": {
+        "type": "boolean",
+        "description": "Clone-status mode: report cloning / done / failed for avatar_id."
       },
       "text": {
         "type": "string",
@@ -85,7 +113,7 @@
       },
       "json_output": {
         "type": "boolean",
-        "description": "Emit a JSON result (generate: { url, generationId }; list: { avatars }; check-status: { status, url }) instead of human-readable output"
+        "description": "Emit a JSON result (generate: { url, generationId }; list: { avatars }; check-status: { status, url }; clone / clone-status: { avatarId, cloneStatus }) instead of human-readable output"
       }
     },
     "required": []
@@ -105,6 +133,12 @@
     "hidden": [
       "json_output",
       "list_avatars",
+      "clone",
+      "clone_url",
+      "clone_type",
+      "thumbnail_url",
+      "wait",
+      "clone_status",
       "mine",
       "name",
       "gender",

package/skills/gen-digital-human/version.json

@@ -2,6 +2,6 @@
   "skillName": "gen-digital-human",
   "repoName": "agent-skill-media-maker",
   "skillId": "336",
-  "version": "V8",
+  "version": "V9",
   "skillDescription": "Digital-human (talking-head) skill: list available avatars, run TTS-based or audio-driven lip-sync, and check the status of pending jobs. Backed by ab-api's digital-human endpoints (Jimeng / HiFly providers).\n\nUse this skill immediately whenever the user asks for any of:\n- Digital human, talking-head video, AI presenter, virtual host\n- \"Make a talking-head video out of this script / this audio file\"\n- List or browse digital-human avatars\n- Use Jimeng or HiFly to drive an avatar\n\nEven without an explicit \"use AI\", any request that turns text or audio into a synthesized presenter video should route here."
 }