🤖 Add truncation: auto to OpenAI Responses API

Enables automatic conversation truncation for OpenAI Responses API to
prevent context overflow errors. When set to 'auto', the API will
automatically drop input items in the middle of the conversation to fit
within the model's context window.

This prevents failures when the conversation history exceeds the
available context size and allows long conversations to continue
seamlessly.

_Generated with `cmux`_

Author: ammario

src/utils/ai/providerOptions.ts

@@ -111,6 +111,7 @@ export function buildProviderOptions(
         parallelToolCalls: true, // Always enable concurrent tool execution
         // TODO: allow this to be configured
         serviceTier: "priority", // Always use priority tier for best performance
+        truncation: "auto", // Automatically truncate conversation to fit context window
         // Conditionally add reasoning configuration
         ...(reasoningEffort && {
           reasoningEffort,