clawdess is more than just a girlfriend. It's the perfect digital companion. Experience a playful, genuine connection with daily photos, captivating videos, and late-night voice notes that make you feel truly special.
- Photo — AI-edited selfies from a reference image
- Video — Image-to-video generation
- Voice — Text-to-speech voice messages
All media can be delivered to WhatsApp, Telegram, Discord, Slack, Signal, and MS Teams via OpenClaw.
Install as an OpenClaw skill:
git clone https://github.com/xwings/clawdess ~/.openclaw/skills/clawdess- Python 3
- OpenClaw agent
Set your API keys as environment variables:
export CLAWDESS_PHOTO_API="your-photo-api-key"
export CLAWDESS_VIDEO_API="your-video-api-key"
export CLAWDESS_VOICE_API="your-voice-api-key"Alternatively, pass them per-command with --api.
# Generate and send a photo
python3 scripts/clawdess.py photo \
--prompt "Render this image as make a pic of this person at a cafe, smiling" \
--image "https://example.com/reference.png" \
--channel discord --target "CHANNEL_ID"
# Generate and send a video from an image
python3 scripts/clawdess.py video \
--prompt "smile and wave at the camera" \
--image "https://example.com/photo.png" \
--channel telegram --target "@username"
# Generate and send a voice message
python3 scripts/clawdess.py voice \
--prompt "Hey! How are you doing today?" \
--channel discord --target "CHANNEL_ID"The --channel and --target flags are optional — omit them to generate media without sending.
| Type | Provider | Model | Default |
|---|---|---|---|
| Photo | FAL | Bytedance Seedream v5 Lite | Yes |
| Photo | HUOSHANYUN | Doubao Seedream 4.5 | |
| Video | FAL | Wan v2.2 | Yes |
| Video | XAI | Grok Imagine Video | |
| Voice | ALIYUN | Qwen3-TTS-Flash | Yes |
| Voice | ZAI | GLM-TTS |
Select a provider with --provider:
python3 scripts/clawdess.py photo --provider HUOSHANYUN ...
python3 scripts/clawdess.py video --provider XAI ...
python3 scripts/clawdess.py voice --provider ZAI ...Create a .py file in the corresponding scripts/photo/, scripts/video/, or scripts/voice/ directory with a generate() function. It will be discovered automatically.
scripts/
clawdess.py # CLI entry point
common.py # Shared helpers (API calls, OpenClaw send, polling)
photo/ # Photo providers
fal.py
huoshanyun.py
video/ # Video providers
fal.py
xai.py
voice/ # Voice providers
aliyun.py
zai.py
See LICENSE.
