Why It Matters
Voice agents are harder to debug than web apps—you can’t just refresh the page. Problems with latency, interruption handling, or tool execution only surface during real conversations. The SDK gives you visibility into what’s happening at every step.Key Capabilities
| Capability | Description |
|---|---|
| Local Testing | Run agents on your machine with the CLI chat interface |
| Logging | Trace events, LLM calls, and tool executions |
| Error Handling | Graceful recovery from failures mid-conversation |
| Performance | Monitor latency and optimize response times |