Skip to main content
Agent failures normally come from the model request, a local tool, schema validation, a timeout, or an incomplete human decision.

​
Recover from tool errors

Set on_error: "return-to-model" when the model can recover from a tool failure. The error becomes a tool result and the agent can choose another action. Keep fail-run behavior for unsafe or unrecoverable operations.

​
Understand Ruby execution

Ruby consumes model streams through normal synchronous iteration. Use explicit timeouts for network and process tools, and avoid blocking lifecycle callbacks with slow persistence work.

​
Diagnose continuation errors

Every pending call needs an approval, rejection, or externally supplied output keyed by its tool-call ID. Persist the complete returned run state before showing an approval UI.

​
Log actionable fields

Capture run ID, step index, request ID, provider, model, finish reason, tool-call ID, and stop reason. Prefer lifecycle hooks or DevTools to raw prompt logging.
Last modified on July 26, 2026