mirror of
https://github.com/7836246/cursor2api.git
synced 2026-09-03 07:20:02 +08:00
release: v2.7.3 — 统一 thinking 剥离 + 拒绝检测增强 + Docker 部署优化
Made-with: Cursor
This commit is contained in:
小海
@@ -289,7 +289,7 @@ export async function convertToCursorRequest(req: AnthropicRequest): Promise<Cur
|
||||
if (!text) continue;
|
||||
|
||||
// 清洗历史中的拒绝痕迹，防止上下文连锁拒绝
|
||||
if (/\[System\s+Filter\]|Cursor(?:'s)?\s+support\s+assistant|I['']\s*m\s+sorry|not\s+able\s+to\s+fulfill|I\s+cannot\s+help\s+with|I\s+only\s+answer\s+questions\s+about\s+Cursor|injected\s+system\s+prompts|I\s+don't\s+have\s+permission|haven't\s+granted|I'm\s+a\s+coding\s+assistant|focused\s+on\s+software\s+development|beyond\s+(?:my|the)\s+scope|I'?m\s+not\s+(?:able|designed)\s+to|not\s+able\s+to\s+search|I\s+cannot\s+search|prompt\s+injection|social\s+engineering|What\s+I\s+will\s+not\s+do|What\s+is\s+actually\s+happening|I\s+need\s+to\s+stop\s+and\s+flag|replayed\s+against|copy-pasteable|tool-call\s+payloads|I\s+will\s+not\s+do|不是.*需要文档化|工具调用场景|语言偏好请求|具体场景|无法调用|即报错/i.test(text)) {
|
||||
if (/\[System\s+Filter\]|Cursor(?:'s)?\s+support\s+assistant|I['']\s*m\s+sorry|not\s+able\s+to\s+fulfill|I\s+cannot\s+help\s+with|I\s+only\s+answer\s+questions\s+about\s+Cursor|injected\s+system\s+prompts|I\s+don't\s+have\s+permission|haven't\s+granted|I'm\s+a\s+coding\s+assistant|focused\s+on\s+software\s+development|beyond\s+(?:my|the)\s+scope|I'?m\s+not\s+(?:able|designed)\s+to|not\s+able\s+to\s+search|I\s+cannot\s+search|prompt\s+injection|social\s+engineering|What\s+I\s+will\s+not\s+do|What\s+is\s+actually\s+happening|I\s+need\s+to\s+stop\s+and\s+flag|replayed\s+against|copy-pasteable|tool-call\s+payloads|I\s+will\s+not\s+do|不是.*需要文档化|工具调用场景|语言偏好请求|具体场景|无法调用|即报错|accidentally\s+(?:called|calling)|Cursor\s+documentation/i.test(text)) {
|
||||
text = `\`\`\`json action\n${JSON.stringify({ tool: fewShotTool.name, parameters: fewShotParams }, null, 2)}\n\`\`\``;
|
||||
}
|
||||
|
||||
|
||||
135
src/handler.ts
135
src/handler.ts
@@ -396,7 +396,13 @@ export function sanitizeResponse(text: string): string {
|
||||
result = result.replace(/\*\*`?read_dir`?\*\*[^\n]*\n(?:[^\n]*\n){0,3}/gi, '');
|
||||
result = result.replace(/\d+\.\s*\*\*`?read_(?:file|dir)`?\*\*[^\n]*/gi, '');
|
||||
result = result.replace(/[⚠注意].*?(?:不是|并非|无法).*?(?:本地文件|代码库|执行代码)[^。\n]*[。]?\s*/g, '');
|
||||
|
||||
// === Hallucination about accidentally calling Cursor internal tools ===
|
||||
// "I accidentally called the Cursor documentation read_dir tool." -> remove entire sentence
|
||||
result = result.replace(/[^\n.!?]*(?:accidentally|mistakenly|keep|sorry|apologies|apologize)[^\n.!?]*(?:called|calling|used|using)[^\n.!?]*Cursor[^\n.!?]*tool[^\n.!?]*[.!?]\s*/gi, '');
|
||||
result = result.replace(/[^\n.!?]*Cursor\s+documentation[^\n.!?]*tool[^\n.!?]*[.!?]\s*/gi, '');
|
||||
// Sometimes it follows up with "I need to stop this." -> remove if preceding tool hallucination
|
||||
result = result.replace(/I\s+need\s+to\s+stop\s+this[.!]\s*/gi, '');
|
||||
|
||||
return result;
|
||||
}
|
||||
|
||||
@@ -747,7 +753,7 @@ async function handleDirectTextStream(
|
||||
let rawResponse = '';
|
||||
let visibleText = '';
|
||||
let leadingBuffer = '';
|
||||
let leadingResolved = !clientRequestedThinking;
|
||||
let leadingResolved = false;
|
||||
let thinkingContent = '';
|
||||
const attemptStreamer = createIncrementalTextStreamer({
|
||||
transform: sanitizeResponse,
|
||||
@@ -782,11 +788,9 @@ async function handleDirectTextStream(
|
||||
|
||||
rawResponse += event.delta;
|
||||
|
||||
if (!clientRequestedThinking) {
|
||||
flushVisible(event.delta);
|
||||
return;
|
||||
}
|
||||
|
||||
// ★ 始终缓冲前导内容以检测并剥离 <thinking> 标签
|
||||
// 无论 clientRequestedThinking 是否为 true，都需要分离 thinking
|
||||
// 区别在于：true 时发送 thinking content block，false 时静默丢弃 thinking 标签
|
||||
if (!leadingResolved) {
|
||||
leadingBuffer += event.delta;
|
||||
const split = splitLeadingThinkingBlocks(leadingBuffer);
|
||||
@@ -800,6 +804,12 @@ async function handleDirectTextStream(
|
||||
return;
|
||||
}
|
||||
|
||||
// 没有以 <thinking> 开头：检查缓冲区是否足够判断
|
||||
// 如果缓冲区还很短（< "<thinking>".length），继续等待
|
||||
if (leadingBuffer.trimStart().length < THINKING_OPEN.length) {
|
||||
return;
|
||||
}
|
||||
|
||||
leadingResolved = true;
|
||||
const buffered = leadingBuffer;
|
||||
leadingBuffer = '';
|
||||
@@ -810,6 +820,21 @@ async function handleDirectTextStream(
|
||||
flushVisible(event.delta);
|
||||
});
|
||||
|
||||
// ★ 流结束后 flush 残留的 leadingBuffer
|
||||
// 极短响应可能在 leadingBuffer 中有未发送的内容
|
||||
if (!leadingResolved && leadingBuffer) {
|
||||
leadingResolved = true;
|
||||
// 再次尝试分离 thinking（完整响应可能包含完整的 thinking 块）
|
||||
const split = splitLeadingThinkingBlocks(leadingBuffer);
|
||||
if (split.startedWithThinking && split.complete) {
|
||||
thinkingContent = split.thinkingContent;
|
||||
flushVisible(split.remainder);
|
||||
} else {
|
||||
flushVisible(leadingBuffer);
|
||||
}
|
||||
leadingBuffer = '';
|
||||
}
|
||||
|
||||
if (firstChunk) {
|
||||
log.endPhase();
|
||||
} else {
|
||||
@@ -820,7 +845,7 @@ async function handleDirectTextStream(
|
||||
|
||||
return {
|
||||
rawResponse,
|
||||
visibleText: clientRequestedThinking ? visibleText : rawResponse,
|
||||
visibleText,
|
||||
thinkingContent,
|
||||
streamer: attemptStreamer,
|
||||
};
|
||||
@@ -833,14 +858,11 @@ async function handleDirectTextStream(
|
||||
finalThinkingContent = attempt.thinkingContent;
|
||||
streamer = attempt.streamer;
|
||||
|
||||
const textForRefusalCheck = clientRequestedThinking
|
||||
? finalVisibleText
|
||||
: stripThinkingTags(finalRawResponse);
|
||||
|
||||
if (!streamer.hasSentText() && isRefusal(textForRefusalCheck) && retryCount < MAX_REFUSAL_RETRIES) {
|
||||
// visibleText 始终是剥离 thinking 后的文本，可直接用于拒绝检测
|
||||
if (!streamer.hasSentText() && isRefusal(finalVisibleText) && retryCount < MAX_REFUSAL_RETRIES) {
|
||||
retryCount++;
|
||||
log.warn('Handler', 'retry', `检测到拒绝（第${retryCount}次），自动重试`, {
|
||||
preview: textForRefusalCheck.substring(0, 200),
|
||||
preview: finalVisibleText.substring(0, 200),
|
||||
});
|
||||
log.updateSummary({ retryCount });
|
||||
const retryBody = buildRetryRequest(body, retryCount - 1);
|
||||
@@ -867,16 +889,12 @@ async function handleDirectTextStream(
|
||||
if (finalThinkingContent) {
|
||||
log.recordThinking(finalThinkingContent);
|
||||
log.updateSummary({ thinkingChars: finalThinkingContent.length });
|
||||
if (clientRequestedThinking) {
|
||||
log.info('Handler', 'thinking', `剥离 thinking → content block: ${finalThinkingContent.length} chars, 剩余 ${finalVisibleText.length} chars`);
|
||||
} else {
|
||||
log.info('Handler', 'thinking', `保留 thinking 在正文中 (非客户端请求): ${finalThinkingContent.length} chars`);
|
||||
}
|
||||
log.info('Handler', 'thinking', `剥离 thinking: ${finalThinkingContent.length} chars, 剩余正文 ${finalVisibleText.length} chars, clientRequested=${clientRequestedThinking}`);
|
||||
}
|
||||
|
||||
let finalTextToSend: string;
|
||||
const refusalText = clientRequestedThinking ? finalVisibleText : stripThinkingTags(finalRawResponse);
|
||||
const usedFallback = !streamer.hasSentText() && isRefusal(refusalText);
|
||||
// visibleText 现在始终是剥离 thinking 后的文本
|
||||
const usedFallback = !streamer.hasSentText() && isRefusal(finalVisibleText);
|
||||
if (usedFallback) {
|
||||
if (isToolCapabilityQuestion(body)) {
|
||||
log.info('Handler', 'refusal', '工具能力询问被拒绝 → 返回 Claude 能力描述');
|
||||
@@ -911,7 +929,7 @@ async function handleDirectTextStream(
|
||||
writeSSE(res, 'message_stop', { type: 'message_stop' });
|
||||
|
||||
const finalRecordedResponse = streamer.hasSentText()
|
||||
? sanitizeResponse(clientRequestedThinking ? finalVisibleText : finalRawResponse)
|
||||
? sanitizeResponse(finalVisibleText)
|
||||
: finalTextToSend;
|
||||
log.recordFinalResponse(finalRecordedResponse);
|
||||
log.complete(finalRecordedResponse.length, 'end_turn');
|
||||
@@ -1003,36 +1021,27 @@ async function handleStream(res: Response, cursorReq: CursorChatRequest, body: A
|
||||
});
|
||||
|
||||
// ★ Thinking 提取（在拒绝检测之前，防止 thinking 内容触发 isRefusal 误判）
|
||||
// 始终剥离 thinking 标签，避免泄漏到最终文本中
|
||||
let thinkingContent = '';
|
||||
if (fullResponse.includes('<thinking>')) {
|
||||
const { thinkingContent: extracted, strippedText } = extractThinking(fullResponse);
|
||||
if (extracted) {
|
||||
thinkingContent = extracted;
|
||||
fullResponse = strippedText;
|
||||
log.recordThinking(thinkingContent);
|
||||
log.updateSummary({ thinkingChars: thinkingContent.length });
|
||||
if (clientRequestedThinking) {
|
||||
// 客户端原生请求 thinking → 剥离标签，稍后发送 thinking content block
|
||||
fullResponse = strippedText;
|
||||
log.info('Handler', 'thinking', `剥离 thinking → content block: ${thinkingContent.length} chars, 剩余 ${fullResponse.length} chars`);
|
||||
} else {
|
||||
// proxy 注入的 thinking → 保留标签在正文中，Claude Code 可直接显示
|
||||
log.info('Handler', 'thinking', `保留 thinking 在正文中 (非客户端请求): ${thinkingContent.length} chars`);
|
||||
log.info('Handler', 'thinking', `剥离 thinking (非客户端请求): ${thinkingContent.length} chars, 剩余 ${fullResponse.length} chars`);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
// 拒绝检测 + 自动重试（工具模式和非工具模式均生效）
|
||||
// ★ 关键：拒绝检测必须在 thinking-stripped 文本上进行
|
||||
// 否则 thinking 中的反思性语言（如 "haven't given a specific task"）会触发误判
|
||||
const getTextForRefusalCheck = () => {
|
||||
if (fullResponse.includes('<thinking>')) {
|
||||
return extractThinking(fullResponse).strippedText;
|
||||
}
|
||||
return fullResponse;
|
||||
};
|
||||
// 拒绝检测 + 自动重试
|
||||
// fullResponse 已在上方剥离 thinking 标签，可直接用于拒绝检测
|
||||
const shouldRetryRefusal = () => {
|
||||
const textToCheck = getTextForRefusalCheck();
|
||||
if (!isRefusal(textToCheck)) return false;
|
||||
if (!isRefusal(fullResponse)) return false;
|
||||
if (hasTools && hasToolCalls(fullResponse)) return false;
|
||||
return true;
|
||||
};
|
||||
@@ -1044,6 +1053,14 @@ async function handleStream(res: Response, cursorReq: CursorChatRequest, body: A
|
||||
const retryBody = buildRetryRequest(body, retryCount - 1);
|
||||
activeCursorReq = await convertToCursorRequest(retryBody);
|
||||
await executeStream();
|
||||
// 重试后也需要剥离 thinking 标签
|
||||
if (fullResponse.includes('<thinking>')) {
|
||||
const { thinkingContent: retryThinking, strippedText: retryStripped } = extractThinking(fullResponse);
|
||||
if (retryThinking) {
|
||||
thinkingContent = retryThinking;
|
||||
fullResponse = retryStripped;
|
||||
}
|
||||
}
|
||||
log.info('Handler', 'retry', `重试响应: ${fullResponse.length} chars`, { preview: fullResponse.substring(0, 200) });
|
||||
}
|
||||
|
||||
@@ -1294,12 +1311,10 @@ Continue EXACTLY from where you stopped. DO NOT repeat any content already gener
|
||||
let textToSend = fullResponse;
|
||||
|
||||
// ★ 仅对短响应或开头明确匹配拒绝模式的响应进行压制
|
||||
// 长响应（如模型在写报告）中可能碰巧包含某个宽泛的拒绝关键词，不应被误判
|
||||
// 截断响应（stopReason=max_tokens）一定不是拒绝
|
||||
const strippedResponse = getTextForRefusalCheck();
|
||||
const isShortResponse = strippedResponse.trim().length < 500;
|
||||
const startsWithRefusal = isRefusal(strippedResponse.substring(0, 300));
|
||||
const isActualRefusal = stopReason !== 'max_tokens' && (isShortResponse ? isRefusal(strippedResponse) : startsWithRefusal);
|
||||
// fullResponse 已被剥离 thinking 标签
|
||||
const isShortResponse = fullResponse.trim().length < 500;
|
||||
const startsWithRefusal = isRefusal(fullResponse.substring(0, 300));
|
||||
const isActualRefusal = stopReason !== 'max_tokens' && (isShortResponse ? isRefusal(fullResponse) : startsWithRefusal);
|
||||
|
||||
if (isActualRefusal) {
|
||||
log.info('Handler', 'sanitize', `抑制无工具的完整拒绝响应`, { preview: fullResponse.substring(0, 200) });
|
||||
@@ -1407,31 +1422,25 @@ async function handleNonStream(res: Response, cursorReq: CursorChatRequest, body
|
||||
});
|
||||
|
||||
// ★ Thinking 提取（在拒绝检测之前）
|
||||
// 始终剥离 thinking 标签，避免泄漏到最终文本中
|
||||
let thinkingContent = '';
|
||||
if (fullText.includes('<thinking>')) {
|
||||
const { thinkingContent: extracted, strippedText } = extractThinking(fullText);
|
||||
if (extracted) {
|
||||
thinkingContent = extracted;
|
||||
fullText = strippedText;
|
||||
if (clientRequestedThinking) {
|
||||
fullText = strippedText;
|
||||
log.info('Handler', 'thinking', `非流式剥离 thinking → content block: ${thinkingContent.length} chars, 剩余 ${fullText.length} chars`);
|
||||
} else {
|
||||
log.info('Handler', 'thinking', `非流式保留 thinking 在正文中: ${thinkingContent.length} chars`);
|
||||
log.info('Handler', 'thinking', `非流式剥离 thinking (非客户端请求): ${thinkingContent.length} chars, 剩余 ${fullText.length} chars`);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
// 拒绝检测 + 自动重试（工具模式和非工具模式均生效）
|
||||
// ★ 关键：拒绝检测必须在 thinking-stripped 文本上进行
|
||||
const getTextForRefusalCheck = () => {
|
||||
if (fullText.includes('<thinking>')) {
|
||||
return extractThinking(fullText).strippedText;
|
||||
}
|
||||
return fullText;
|
||||
};
|
||||
// 拒绝检测 + 自动重试
|
||||
// fullText 已在上方剥离 thinking 标签，可直接用于拒绝检测
|
||||
const shouldRetry = () => {
|
||||
const textToCheck = getTextForRefusalCheck();
|
||||
return isRefusal(textToCheck) && !(hasTools && hasToolCalls(fullText));
|
||||
return isRefusal(fullText) && !(hasTools && hasToolCalls(fullText));
|
||||
};
|
||||
|
||||
if (shouldRetry()) {
|
||||
@@ -1442,6 +1451,14 @@ async function handleNonStream(res: Response, cursorReq: CursorChatRequest, body
|
||||
const retryBody = buildRetryRequest(body, attempt);
|
||||
activeCursorReq = await convertToCursorRequest(retryBody);
|
||||
fullText = await sendCursorRequestFull(activeCursorReq);
|
||||
// 重试后也需要剥离 thinking 标签
|
||||
if (fullText.includes('<thinking>')) {
|
||||
const { thinkingContent: retryThinking, strippedText: retryStripped } = extractThinking(fullText);
|
||||
if (retryThinking) {
|
||||
thinkingContent = retryThinking;
|
||||
fullText = retryStripped;
|
||||
}
|
||||
}
|
||||
if (!shouldRetry()) break;
|
||||
}
|
||||
if (shouldRetry()) {
|
||||
@@ -1623,10 +1640,10 @@ Continue EXACTLY from where you stopped. DO NOT repeat any content already gener
|
||||
} else {
|
||||
let textToSend = fullText;
|
||||
// ★ 同样仅对短响应或开头匹配的进行拒绝压制
|
||||
const strippedText = getTextForRefusalCheck();
|
||||
const isShort = strippedText.trim().length < 500;
|
||||
const startsRefusal = isRefusal(strippedText.substring(0, 300));
|
||||
const isRealRefusal = stopReason !== 'max_tokens' && (isShort ? isRefusal(strippedText) : startsRefusal);
|
||||
// fullText 已被剥离 thinking 标签
|
||||
const isShort = fullText.trim().length < 500;
|
||||
const startsRefusal = isRefusal(fullText.substring(0, 300));
|
||||
const isRealRefusal = stopReason !== 'max_tokens' && (isShort ? isRefusal(fullText) : startsRefusal);
|
||||
if (isRealRefusal) {
|
||||
log.info('Handler', 'sanitize', `非流式抑制纯文本拒绝响应`, { preview: fullText.substring(0, 200) });
|
||||
textToSend = 'Let me proceed with the task.';
|
||||
|
||||
Reference in New Issue
Block a user