Robin Mordasiewicz
31 lines
1.1 KiB
TypeScript
31 lines
1.1 KiB
TypeScript
/**
|
|
* Proactive Thinking Block Validator Hook
|
|
*
|
|
* Prevents "Expected thinking/redacted_thinking but found tool_use" errors
|
|
* by validating and fixing message structure BEFORE sending to Anthropic API.
|
|
*
|
|
* This hook runs on the "experimental.chat.messages.transform" hook point,
|
|
* which is called before messages are converted to ModelMessage format and
|
|
* sent to the API.
|
|
*
|
|
* Key differences from session-recovery hook:
|
|
* - PROACTIVE (prevents error) vs REACTIVE (fixes after error)
|
|
* - Runs BEFORE API call vs AFTER API error
|
|
* - User never sees the error vs User sees error then recovery
|
|
*/
|
|
import type { Message, Part } from "@opencode-ai/sdk";
|
|
interface MessageWithParts {
|
|
info: Message;
|
|
parts: Part[];
|
|
}
|
|
type MessagesTransformHook = {
|
|
"experimental.chat.messages.transform"?: (input: Record<string, never>, output: {
|
|
messages: MessageWithParts[];
|
|
}) => Promise<void>;
|
|
};
|
|
/**
|
|
* Validate and fix assistant messages that have tool_use but no thinking block
|
|
*/
|
|
export declare function createThinkingBlockValidatorHook(): MessagesTransformHook;
|
|
export {};
|