Skip to main content

avatar-create

Use to build a reusable custom avatar identity in Creatify from the user's own media — a talking video (bring-your-own-avatar, needs a consent clip), a single face photo, or a text description. Covers avatar_create_video / avatar_create_image / avatar_create_text, avatar_choose_photo, avatar_create_get, and the upload widget. This makes an avatar IDENTITY, not a talking video.

ソース情報

リポジトリ
creatify-ai/Creatify-agent-plugin
ソースの最終更新活動
2026年7月16日 07:20
検出された SKILL.md の言語
英語
スター
1
フォーク
0

インストール方法

デフォルトでは、最初にソースを確認する Prompt が選択されています。直接コマンドに切り替えるか、ローカルコピーをダウンロードすることもできます。

ソースファイルを確認

インストールを決める前に、SKILL.md と SkillsMP に表示されている付属ファイルをお読みください。

SKILL.md を表示中

SKILL.md
ソースの指示 · 読み取り専用プレビュー
name
avatar-create
description
Use to build a reusable custom avatar identity in Creatify from the user's own media — a talking video (bring-your-own-avatar, needs a consent clip), a single face photo, or a text description. Covers avatar_create_video / avatar_create_image / avatar_create_text, avatar_choose_photo, avatar_create_get, and the upload widget. This makes an avatar IDENTITY, not a talking video.
# Create a Custom Avatar ## Purpose Build a **reusable avatar identity** the user can then use to render talking videos. This is distinct from making a talking video: once the avatar is ready it appears in `avatar_list`, and you generate speech with it via `avatar-video` (Aurora). ## Three ways to create - **From a talking video (BYOA)** — `avatar_create_video`. Needs the footage, a **consent clip**, and a name. If the consent clip or name is missing, open the video wizard with `avatar_create_video_widget` (two upload slots: footage + consent) and let the user provide them. Consent is required — do not bypass it. - **From one face photo** — `avatar_create_image`. A reusable avatar from a single face photo (~5–10 min). - **From a text description (synthetic)** — `avatar_create_text` generates candidate faces; then submit the chosen candidate with `avatar_choose_photo` to build the avatar. Read each tool's schema for exact params before calling. ## Flow 1. Pick the tier that matches what the user has (video / photo / text). 2. Start creation; for the video tier, ensure footage + consent + name (widget if missing). 3. Poll `avatar_create_get` until done. The `avatar_id` populates when ready and the avatar also shows up in `avatar_list`. 4. If the user's goal was a talking video, hand off to `avatar-video` (Aurora) using the new `avatar_id`. ## Async & Verify Creation is long-running. Poll `avatar_create_get`; only report success once an `avatar_id` is returned. See `verification` and `known-errors`.
GitHubで見る