Documentation
Build with taste
Everything you need to capture, version, and operationalize human aesthetic preferences.
Quickstart
Get started in 3 steps
1
Create a profile
bash
curl -X POST https://api.commandagi.com/v1/profiles \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{"name": "my-taste-profile"}'2
Add preferences
bash
curl -X POST https://api.commandagi.com/v1/profiles/PROFILE_ID/labels \
-H "Authorization: Bearer YOUR_API_KEY" \
-d '{"frame_url": "https://...", "label": "good"}'3
Evaluate content
bash
curl -X POST https://api.commandagi.com/v1/profiles/PROFILE_ID/eval \
-H "Authorization: Bearer YOUR_API_KEY" \
-d '{"frame_url": "https://your-image.jpg"}'
# Response: {"score": 0.847, "confidence": 0.92}Documentation
Explore the docs
Getting Started
Create your first taste profile in minutes
API Reference
Complete reference for all endpoints
Marketplace
Scale annotation with our network
SDKs
Native libraries
Use your favorite language
JavaScript
npm install @commandagi/sdkPython
pip install commandagiGo
go get github.com/commandagi/goREST
curl https://api.commandagi.com/v1/API Design
Clean, predictable responses
JSON responses with consistent structure across all endpoints
json
{
"id": "prof_abc123",
"name": "cinematic-taste",
"version": "v3",
"branch": "main",
"created_at": "2024-01-15T10:30:00Z",
"stats": {
"total_labels": 1247,
"total_comparisons": 892,
"dimensions": ["composition", "color", "mood"]
},
"scoring": {
"model": "bradley-terry",
"confidence": 0.94
}
}Get Started
Ready to build?
Get your API key and start building taste-aware applications.