AIF-Execution
Capability Router & Execution Engine β Dispatches tasks to the optimal platform adapter with cost/latency awareness and automatic fallback.
16 Platform Adapters
| Adapter | Type | Status |
|---|---|---|
| OpenClaw | Agent Runtime | β Active |
| CrewAI | Multi-Agent | β Active |
| AgentOS | Agent OS | β Active |
| LangGraph | Workflow | β Active |
| OpenAI SDK | Provider | β Active |
| AutoGen | Multi-Agent | β Active |
| Google ADK | Agent SDK | β Active |
| Semantic Kernel | Agent SDK | β Active |
| Claude Code | Coding Agent | β Active |
| Codex CLI | Coding Agent | β Active |
| Goose | Coding Agent | β Active |
| A2A Protocol | Federation | β Active |
| MCP Protocol | Tool Bridge | β Active |
| OpenTelemetry | Observability | β Active |
| Kubernetes Jobs | Container | β Active |
| Local Runtime | Process | β Active |
Routing Logic
- Parse task requirements (model type, latency budget, cost cap)
- Query provider registry for compatible adapters
- Score by health, latency, and cost profile
- Select optimal adapter; configure fallback chain
- Dispatch and monitor execution
Repository
Inference Providers NEW
This model isn't deployed by any Inference Provider. π Ask for provider support