Expand description
AI model definitions and routing.
Defines model roles (Thinking, Coding, Task, Embedding, Auditor) and provides the foundation for multi-model orchestration.
Modules§
- cli_
provider - CLI Provider — use authenticated CLI tools (claude, gemini, auggie) as AI backends.
- ollama
- Ollama local model provider — for Embedding and Auditor roles.
- openrouter
- OpenRouter provider — access multiple models (Gemini Flash, etc.) via unified API.
- provider
- Model provider — Anthropic API client for AI-powered workflow steps.
- router
- Model Router — selects the appropriate AI model based on step type and role.
- schema
- Structured output schemas for AI model tool calling.
Structs§
- Model
Config - Configuration for a specific model.