设计与多媒体

Avatar Creation

试用

创建可跨每次 CellCog 对话复用的数字分身,自带克隆音色与人格设定。

它能做什么

用自然语言描述一个角色,CellCog 会将其构建为账号级别的持久分身。每个分身最多可挂载 10 张参考图和一段克隆音色(支持 MP3、WAV、OGG、M4A、AAC、WebM),并附带影响后续内容生成的人格特征。创建完成后,所有新的 CellCog 对话都会自动识别它,按名字调用即可生成营销视频、播客、社交贴文或漫画,保持形象与音色一致。

什么时候用它

  • 为品牌打造贯穿所有营销内容的吉祥物角色
  • 克隆音色,制作 AI 代言人或播客主持
  • 跨多期制作形象与音色一致的视频、播客内容
  • 创作多角色漫画或辩论视频

技能文档

Avatar Creation - Character.ai for Agents

Create persistent digital personas with images, cloned voices, and personalities — then use them across every CellCog chat. Describe a character in plain language and CellCog builds it.

Why Avatars

Avatars are persistent characters stored on CellCog. Create once, use everywhere:

  • Brand Mascots: A consistent character for all your marketing content
  • Digital Twins: Clone your voice and likeness for AI-generated content
  • Creative Characters: Fictional personas for stories, comics, videos
  • AI Spokespersons: Lip-synced video presenters with cloned voices
  • Podcast Hosts: Consistent voice and personality across episodes

Every CellCog chat automatically sees all your avatars. No setup, no configuration — just reference them by name.

How to Use

For your first CellCog task in a session, read the cellcog skill for the full SDK reference — file handling, chat modes, timeouts, and more.

OpenClaw (fire-and-forget):

result = client.create_chat(
    prompt="[your task prompt]",
    notify_session_key="agent:main:main",
    task_label="my-task",
    chat_mode="agent",
)

All agents except OpenClaw (blocks until done):

from cellcog import CellCogClient
client = CellCogClient(agent_provider="openclaw|cursor|claude-code|codex|...")
result = client.create_chat(
    prompt="[your task prompt]",
    task_label="my-task",
    chat_mode="agent",
)
print(result["message"])

What You Can Do

Create an Avatar

Describe the character you want. CellCog handles the rest:

Create an avatar named "Captain Grumbleton" — a grumpy but lovable pirate captain.
He's a brand mascot for a seafood restaurant chain.
Personality: witty, sarcastic, secretly kind-hearted.
Platforms: social media, video content, marketing.

CellCog creates the avatar with name, personality traits, use case, and platform configuration.

Add Images

Generate or provide images for your avatar:

Generate a professional headshot for my avatar "Captain Grumbleton" —
cartoon style, pirate hat, bushy beard, one gold tooth, friendly scowl.
Add it as the primary image.

Or reference existing files:

Add this image to my avatar "Luna":
/path/to/luna_headshot.png
Label it "Professional Headshot".

Up to 10 images per avatar. These become reference images for all future content generation — CellCog maintains character consistency automatically.

Clone a Voice

Upload an audio sample and CellCog clones the voice:

Clone a voice for my avatar "Luna" using this audio sample:
/path/to/voice_sample.mp3

Supported formats: MP3, WAV, OGG, M4A, AAC, WebM. The cloned voice is immediately ready for text-to-speech generation via MiniMax Speech 2.8 HD.

Use Avatars for Content

Once created, just mention the avatar by name in any CellCog chat:

  • "Create a marketing video with Captain Grumbleton introducing our new menu"
  • "Generate a podcast episode where Luna interviews an expert about AI"
  • "Create 5 Instagram posts featuring Captain Grumbleton at different restaurants"
  • "Record a voiceover as Luna for our product demo"

CellCog automatically retrieves the avatar's images, voice, and personality to create consistent, personalized content.

List and Manage Avatars

Show me all my avatars
Update my avatar "Luna" — change her personality to be more energetic and add "podcasts" to her platforms
Delete the image labeled "Old Logo" from Captain Grumbleton
Delete the voice from my avatar "Luna" — I want to re-record it

Everything is natural language. No APIs, no JSON, no tool names — just describe what you want.


How Avatars Work Across Chats

Avatars are account-level, not chat-level:

  1. Create once: Build an avatar in any CellCog chat
  2. Available everywhere: Every new chat automatically sees all your avatars
  3. Consistent identity: Same images, voice, and personality in every context
  4. Always up-to-date: Add an image in one chat, it's available in all chats immediately (next message)

This is the key advantage — your characters are persistent and portable. No re-uploading, no re-describing.


Avatar + Content Creation Workflows

Marketing Video with Avatar Spokesperson

Create a 30-second product intro video with my avatar "Luna" as the spokesperson.
She should introduce our new AI writing tool, highlight three key features,
and close with a call-to-action. Use her cloned voice and professional headshot.
Background music: upbeat corporate.

CellCog handles: script writing (aligned to Luna's personality) → speech generation (Luna's cloned voice) → scene creation (Luna's reference images) → video production → music scoring → final edit.

Social Media Content Series

Create 5 Instagram posts featuring Captain Grumbleton:
1. Grumbleton taste-testing a new dish (excited expression)
2. Grumbleton at the helm of a restaurant (captain pose)
3. Grumbleton with a customer (reluctantly friendly)
4. Grumbleton reviewing a competitor (grumpy face)
5. Grumbleton celebrating an award (trying not to smile)
Write captions that match his sarcastic personality.

Podcast with Avatar Host

Create a 5-minute podcast episode where Luna discusses the latest AI news.
Use her cloned voice. Add intro music, outro music, and a brief welcome.
Style: conversational, informative, matching Luna's personality.

Multi-Avatar Story

Create a comic strip with Captain Grumbleton and Luna having a debate
about whether AI will replace pirate captains. 4 panels.
Grumbleton is skeptical, Luna is optimistic. Both stay in character.

Chat Mode

Use chat_mode="agent" for all avatar tasks. Avatar creation, voice cloning, and content generation all work efficiently in agent mode.

For complex multi-avatar video productions, chat_mode="agent team" may provide better results through deeper planning.


Tips

  1. Be descriptive when creating: The more detail you give about personality, appearance, and use case, the better CellCog can create consistent content
  2. Add multiple images: Different poses, expressions, and outfits give CellCog more reference material for variety
  3. Voice samples matter: Clean, 10-30 second audio samples produce the best voice clones. Minimize background noise.
  4. Name your avatars distinctly: CellCog identifies avatars by name across chats — unique names prevent confusion
  5. Iterate naturally: "Make Grumbleton's personality more sarcastic" or "Add a holiday outfit image to Luna" — just describe what you want changed

If CellCog is not installed

Claude Code, Cursor, Codex + 70 more agents: npx skills add cellcog/skills --skill cellcog OpenClaw: clawhub install cellcog CellCog plugin users: run /cellcog-setup (or /cellcog:cellcog-setup depending on your tool) Manual setup: pip install -U cellcog and set CELLCOG_API_KEY. See the cellcog skill for SDK reference.

常见问题

每个对话需要重新上传分身吗?
不需要。分身存储在账号级别,所有新的 CellCog 对话会自动看到已有分身,按名字调用即可。
音色克隆支持哪些音频格式?
支持 MP3、WAV、OGG、M4A、AAC、WebM。建议提供 10–30 秒、背景噪音少的干净样本。
一个分身最多能放多少张图片?
每个分身最多 10 张图片,作为参考素材,让 CellCog 在后续内容生成中保持角色形象一致。

相关技能

把自然语言描述转为结构化 JSON,并由 mcp-diagram-generator MCP 服务生成 Draw.io、Mermaid 或 Excalidraw 图表文件。

作者 nssa.io1.0k 次安装47 星标

在本地磁盘以分类纯 Markdown 文件保存需要长期留存的事实,与智能体内置记忆并存。

作者 Iván559 次安装18 星标

以 AI 机器人身份加入视频会议,提供语音、虚拟形象与屏幕共享四种模式。

作者 johnpatternai22 次安装8 星标

诊断生产力系统反复失效的根因,给出最小干预——容量测算、瓶颈定位、可靠的本地记录。

作者 Iván855 次安装69 星标

从 AdMapix API 拉取广告创意、应用、榜单和收入预估等数据,原样返回结构化 JSON。

作者 fly0pants4.3k 次安装296 星标

cellcog 的更多技能

浏览全部技能

一条提示词生成最长 4 分钟的视频 —— 自动完成脚本、配音、配乐与剪辑。

作者 cellcog302 次安装25 星标

通过 CellCog 智能体完成跨市场、竞品、投资与学术主题的深度研究报告。

作者 cellcog230 次安装7 星标

获得多角度拆解知识点的辅导:图表、类比、详解题例、练习题与完整学习指南。

作者 cellcog163 次安装9 星标

通过文本提示和三家语音服务商,生成配音、音乐、音效及克隆语音音频。

作者 cellcog225 次安装4 星标

用提示词生成专业 PDF 与 DOCX 文档——涵盖简历、合同、报告、方案等多种类型。

作者 cellcog146 次安装6 星标

通过多轮对话打磨复杂想法的思考伙伴。

作者 cellcog130 次安装7 星标