veya2ztn

kameo

设计创意v1.0.1
作者:veya2ztn更新于 2026-05-18共 2 个版本 · 最新版本发布于 2026-02-04@clawhub_veya2ztn/kameo
图片生成视频处理#Video Generation#Sync

据 AlphaMoat 统计,截至 2026-09,kameo 累计下载 3,063 次、获 5 个星标,在全部 5,078 个 AI 技能中排名第 3,687 位,在「设计创意」分类(512 个)中排名第 347 位。

累计安装
223
累计下载
3,063
星标
5
全站排名
#3,687
全站前 73% · 「设计创意」前 68%
平台累计安装累计下载星标评论
ClawHub ↗912,81051
SkillHub ↗2233,0635—

各平台的计数口径不同,统计时间不一致。

这个技能是做什么的

向 Kameo 提交 JPEG/PNG 图像及对白或场景提示词,生成说话头像视频,并返回视频地址。需要 API 密钥和可用积分;处理耗时受画幅与队列影响。CDN 地址可能限时或要求鉴权,使用 curl 下载可能返回 403。

资料发布于 2026-10-26

可以用来做什么

  • 制作角色介绍或讲解视频

什么时候用

  • 已有肖像和对白,需要生成说话头像片段时

示例提示词

以下是根据技能文档生成的示例,不代表实际运行结果。

用这张头像和对白生成讲解视频

使用条件

依据技能文档中的声明整理,未经实际安装或运行验证。

鉴权说明
配置 KAMEO_API_KEY 或 ~/.config/kameo/credentials.json 中的密钥。
配置要求
  • 来源标示每段视频消耗 3 积分,每分钟最多生成 10 次;余额不足会报错。

月度数据趋势

累计下载3,063
1,9022,5263,1492026-03 · 累计下载 1,9882026-05 · 累计下载 2,3212026-06 · 累计下载 2,4632026-07 · 累计下载 2,6702026-08 · 累计下载 2,7372026-09 · 累计下载 3,06326-0326-09
星标5
4562026-03 · 星标 52026-05 · 星标 52026-06 · 星标 52026-07 · 星标 52026-08 · 星标 52026-09 · 星标 526-0326-09

2026-09 起统计方法调整:同名技能按作者分别统计,并覆盖全部技能。图中虚线为调整前的月份,与之后的数据不直接比较,也不计算跨越调整的增长。

月份累计下载环比星标
2026-093,063—5
2026-082,737+2.5%5
2026-072,670+8.4%5
2026-062,463+6.1%5
2026-052,321—5
2026-031,988—5

SKILL.md 原文

以下是作者发布的原始文件,按纯文本显示,本站未做修改或审核;其中的链接不可点击,命令请核实后再运行。抓取于 2026-10-23。

5.7 KB · ClawHub v1.0.1
---
name: kameo
description: Generate expressive talking-head videos from static images using Kameo AI. Converts static avatars/portraits into dynamic 5-second videos with realistic facial expressions, lip-sync, and motion. Use when you need to bring static images to life, create AI character videos, demonstrate visual communication, or generate talking avatars from photos.
---

# Kameo AI - Talking Head Video Generation

Transform static images into expressive talking-head videos with realistic motion and lip-sync.

## Quick Start
展开全部(共 205 行)

```bash
scripts/generate_video.sh <image_path> <prompt> [output_file]
```

**Example:**
```bash
scripts/generate_video.sh avatar.jpg "Hello, I am an AI assistant" output.mp4
```

## What It Does

- Takes a static image (portrait/avatar)
- Adds realistic facial motion, expressions, and lip-sync based on your prompt
- Generates 5-second video in 9:16, 16:9, or 1:1 aspect ratio
- Returns CDN URL instantly (processing ~10-30 seconds)

## Authentication

Set your Kameo API key:
```bash
export KAMEO_API_KEY="kam_I3rdx43IymFNbfBw1c0ZbSc7o3aUfQgz8cljZA6T7fs"
```

Or store in `~/.config/kameo/credentials.json`:
```json
{
  "api_key": "kam_I3rdx43IymFNbfBw1c0ZbSc7o3aUfQgz8cljZA6T7fs"
}
```

**Getting an API Key:**

1. Register at kameo.chat (requires email verification)
2. Login to get JWT token
3. Create API key via `/api/public/keys` endpoint
4. Or use the registration helper: `scripts/register.sh`

## Prompt Engineering

### Basic Prompts (Simple)

Just the dialogue:
```
"Hello, I'm here to help you today"
"こんにちは、私はガッキーです。愛してます。"
```

Works but results are generic.

### Enhanced Prompts (Recommended)

**Format:**
```
[Detailed scene/environment], [person's complete appearance and expression], speaking in [tone], "[DIALOGUE]". [Camera and lighting details].
```

**Example:**
```
In a bright outdoor winter setting with soft, overcast daylight, a young woman with long dark hair wearing a white knitted winter hat with ear flaps and a colorful patterned sweater stands centered in frame. She looks directly into the camera with a warm, genuine smile, her eyes crinkling with joy, speaking in a cheerful, affectionate tone, "こんにちは、私はガッキーです。愛してます。" The scene is captured in a medium close-up shot, framed at eye level. The lighting is natural and diffused from above, creating soft, even illumination.
```

**Why Enhanced Prompts Matter:**
- Better facial expressions matching the scene context
- More natural motion and gestures
- Improved lip-sync quality
- Contextual emotional delivery

### Prompt Enhancement Workflow

For best results, use vision AI to analyze the image first:

1. Feed the image to a vision model (Gemini, GPT-4V, Claude)
2. Ask it to describe the scene in cinematic detail
3. Insert your dialogue into the description
4. Use the enhanced prompt for Kameo

**See:** `scripts/enhance_prompt.sh` for automated enhancement.

## API Details

**Base URL:** `https://api.kameo.chat/api/public`

### Generate Video

```bash
curl -X POST https://api.kameo.chat/api/public/generate \
  -H "X-API-Key: kam_I3rdx43IymFNbfBw1c0ZbSc7o3aUfQgz8cljZA6T7fs" \
  -H "Content-Type: application/json" \
  -d '{
    "image_base64": "<base64_encoded_image>",
    "prompt": "Your detailed prompt here",
    "seconds": 5,
    "aspect_ratio": "9:16"
  }'
```

**Parameters:**
- `image_base64` (required): Base64-encoded JPEG/PNG
- `prompt` (required): Dialogue and/or scene description
- `seconds` (optional): 5 (default) or 10
- `aspect_ratio` (optional): "9:16" (default), "16:9", or "1:1"

**Response:**
```json
{
  "job_id": "uuid",
  "status": "completed",
  "video_url": "https://cdn.kameo.chat/videos/{uuid}.mp4",
  "duration_seconds": 5,
  "processing_time_ms": 15000
}
```

### Check Credits

```bash
curl -H "X-API-Key: kam_..." \
  https://api.kameo.chat/api/public/credits
```

**Response:**
```json
{
  "permanent_credits": 294,
  "subscription_credits": 0,
  "total_available": 294
}
```

### Pricing

```bash
curl https://api.kameo.chat/api/public/pricing
```

**Cost:** 3 credits per video

## Performance

- **Processing time:** 8-35 seconds (depends on aspect ratio and queue)
- **9:16 (portrait):** ~30-35s
- **16:9 (landscape):** ~15-20s  
- **1:1 (square):** ~10-15s

## Best Practices

1. **Optimize image size** - Resize large images before encoding (saves bandwidth, faster upload)
   ```bash
   ffmpeg -i large.jpg -vf scale=720:-1 optimized.jpg
   ```

2. **Use descriptive prompts** - Enhanced prompts = better results

3. **Choose aspect ratio wisely**
   - 9:16: Mobile/social media (TikTok, Instagram Stories)
   - 16:9: Desktop/YouTube
   - 1:1: Profile pictures, square posts

4. **Monitor credits** - Check balance with `scripts/check_credits.sh`

## Limitations

- **CDN access:** Video URLs may have time-limited access or require authentication
- **Download:** Videos may return 403 when downloaded via curl (use browser or authenticated session)
- **Rate limits:** 10 generations per minute

## Troubleshooting

**"401 Unauthorized"**
- Check your API key is set correctly
- Verify key hasn't been revoked

**"402 Insufficient credits"**
- Check credit balance: `scripts/check_credits.sh`
- Need to add credits at kameo.chat

**"Timeout errors"**
- 9:16 videos take longer (~30s)
- Increase timeout in scripts
- Retry if server is busy

**"403 when downloading video"**
- CDN URLs may be time-limited
- Try accessing in browser immediately after generation
- Or save the base64 response if available

## Use Cases

- **AI character videos** - Bring bot avatars to life
- **Social media content** - Dynamic profile videos
- **Demos and presentations** - Talking product demos
- **Educational content** - Video tutorials with AI presenters
- **Multilingual content** - Same avatar speaking different languages

「设计创意」分类下的热门技能

架构图一键生成设计创意

使用 ContextWeave 生成或修改架构图、流程图、思维导图和复杂信息图。适用于需要从代码、文件或自然语言中提取结构与关系并实际产出 CW、SVG、HTML、可编辑 PPTX 或原生 VSDX 的请求;未特别说明时,PPTX 与 Visio 使用原生形状和连接器导出。不适用于统计图表、手绘插画或像素级排版。

↓ 28.2万★ 250user_bddf3fe6
Nano Banana Pro设计创意

使用 Nano Banana Pro (Gemini 3 Pro Image) 生成或编辑图像。支持文生图、图生图及 1K/2K/4K 分辨率,适用于图像创建、修改及编辑请求,使用 --input-image 指定输入图像。

↓ 14.1万★ 473Peter Steinberger
Openai Whisper设计创意

使用 Whisper CLI 进行本地语音转文字(无需 API 密钥)

↓ 11.4万★ 359Peter Steinberger
diagram-generator设计创意

使用 mcp-diagram-generator MCP 服务器生成和编辑图表。适用于新建图表、编辑现有 .drawio/.mmd/.excalidraw 文件、网络拓扑、架构图、流程图、泳道图、时序图、类图、ER 图及 Excalidraw 白板工作。当用户要求绘图时,始终使用此技能。

↓ 9.8万★ 73nssa.io
UI/UX Pro Max设计创意

面向界面设计的思路与落地建议技能,覆盖布局、交互流程、设计令牌与前端改动。

↓ 8.5万★ 290xobi667
AI Video Generation设计创意

使用 Sora 2、Veo 3、Seedance、Runway 及现代 API,通过可靠的提示与渲染工作流创建 AI 视频。

↓ 7.2万★ 10Iván
Video Frames设计创意

用 ffmpeg 从视频里取出画面帧,或生成便于快速查看的缩略图。

↓ 6.8万★ 145Peter Steinberger
Local Whisper设计创意

用本地 Whisper 把音频文件转成文字,模型下载完成后可完全离线运行。

↓ 6.2万★ 19araa47

数据月份:2026-09 · 下载/星标/安装数聚合自公开技能注册库(ClawHub、SkillHub),按月更新。口径说明见数据方法。