Configuration
Configure your SwiftClaw agents and environment
Configuration
Learn how to configure SwiftClaw agents, models, memory, and deployment settings.
Configuration File
SwiftClaw uses swiftclaw.json for agent configuration:
{
"name": "my-agent",
"version": "1.0.0",
"runtime": "python3.11",
"model": {
"primary": "gpt-4",
"fallbacks": ["claude-3-sonnet"]
},
"memory": {
"type": "persistent",
"ttl": "30d"
},
"triggers": [
{
"type": "api",
"path": "/chat"
}
]
}Model Configuration
Configure AI models and fallbacks:
{
"model": {
"primary": "gpt-4",
"fallbacks": ["claude-3-sonnet", "gemini-pro"],
"temperature": 0.7,
"maxTokens": 2000,
"topP": 0.9
}
}Memory Configuration
Set up memory persistence:
{
"memory": {
"type": "persistent",
"scope": "per-user",
"ttl": "30d",
"maxSize": "100MB"
}
}Environment Variables
Store sensitive data in .env:
OPENAI_API_KEY=sk-...
ANTHROPIC_API_KEY=sk-ant-...
SWIFTCLAW_API_KEY=your-api-keyNever commit .env files to version control.
Next Steps
- CLI Reference - Learn CLI commands
- Deployment - Deploy your agent
How is this guide ?
Last updated on