From f0caa04f60de7380ab8284ce62cde4e24163e4e0 Mon Sep 17 00:00:00 2001
From: hobostay <hobostay@users.noreply.github.com>
Date: Thu, 12 Mar 2026 16:57:23 +0800
Subject: [PATCH] fix: strip leaked Thinking/Final tags from OpenAI-compatible
 responses

Some OpenAI-compatible providers (e.g., MiniMax) may return chain-of-thought
content mixed into the response using tags like:
- </think>...</think>
- <thinking>...</thinking>
- <thought>...</thought>
- <reasoning>...</reasoning>
- <final>...</final>

These tags can leak internal reasoning text to end users. This commit adds
a sanitizer that:
1. Handles escaped Unicode forms (\u003c/\u003e)
2. Removes thinking blocks entirely (tags + content)
3. Removes <final> tag wrappers while preserving the final content
4. Cleans up leftover whitespace-only lines

Fixes #1235

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
---
 pkg/providers/openai_compat/provider.go | 56 ++++++++++++++++++++++++-
 1 file changed, 55 insertions(+), 1 deletion(-)

diff --git a/pkg/providers/openai_compat/provider.go b/pkg/providers/openai_compat/provider.go
index f97bf3acd..f887202f3 100644
--- a/pkg/providers/openai_compat/provider.go
+++ b/pkg/providers/openai_compat/provider.go
@@ -10,6 +10,7 @@ import (
 	"log"
 	"net/http"
 	"net/url"
+	"regexp"
 	"strings"
 	"time"
 
@@ -346,8 +347,11 @@ func parseResponse(body io.Reader) (*LLMResponse, error) {
 		toolCalls = append(toolCalls, toolCall)
 	}
 
+	// Strip any leaked thinking/reasoning tags from content
+	cleanedContent := stripThinkingTags(choice.Message.Content)
+
 	return &LLMResponse{
-		Content:          choice.Message.Content,
+		Content:          cleanedContent,
 		ReasoningContent: choice.Message.ReasoningContent,
 		Reasoning:        choice.Message.Reasoning,
 		ReasoningDetails: choice.Message.ReasoningDetails,
@@ -518,3 +522,53 @@ func supportsPromptCacheKey(apiBase string) bool {
 	host := u.Hostname()
 	return host == "api.openai.com" || strings.HasSuffix(host, ".openai.azure.com")
 }
+
+// stripThinkingTags removes chain-of-thought tags from content that may be
+// leaked by some OpenAI-compatible providers (e.g., MiniMax-style responses).
+// Handles both normal tags and escaped Unicode forms.
+// Tags removed: <think>, <thinking>, <thought>, <reasoning>, <final>
+func stripThinkingTags(content string) string {
+	if content == "" {
+		return content
+	}
+
+	// First, handle escaped forms like \u003cthink\u003e...\u003c/think\u003e
+	// These appear as literal backslash-u-sequences in the string
+	content = strings.ReplaceAll(content, `\u003c`, "<")
+	content = strings.ReplaceAll(content, `\u003e`, ">")
+	content = strings.ReplaceAll(content, `\\u003c`, "<")
+	content = strings.ReplaceAll(content, `\\u003e`, ">")
+
+	// Pattern to match thinking/reasoning tags (with optional inner content)
+	// Matches: <tag>content</tag>, <tag />, <tag></tag>
+	patterns := []string{
+		`(?i)<think\b[^>]*>.*?</think>`,   // <think>...</think>
+		`(?i)<thinking\b[^>]*>.*?</thinking>`, // <thinking>...</thinking>
+		`(?i)<thought\b[^>]*>.*?</thought>`,   // <thought>...</thought>
+		`(?i)<reasoning\b[^>]*>.*?</reasoning>`, // <reasoning>...</reasoning>
+		`(?i)<final\b[^>*>(?:/>|>.*?</final>)`, // <final/> or <final>...</final> (remove tag, keep content)
+	}
+
+	for _, pattern := range patterns {
+		re := regexp.MustCompile(pattern)
+		if strings.HasPrefix(pattern, `(?i)<final`) {
+			// For <final> tags, remove the tags but keep the inner content
+			content = re.ReplaceAllString(content, "$1")
+		} else {
+			// For thinking tags, remove both tags and content
+			content = re.ReplaceAllString(content, "")
+		}
+	}
+
+	// Clean up any whitespace-only lines left after removing blocks
+	lines := strings.Split(content, "\n")
+	var cleaned []string
+	for _, line := range lines {
+		trimmed := strings.TrimSpace(line)
+		if trimmed != "" {
+			cleaned = append(cleaned, line)
+		}
+	}
+
+	return strings.Join(cleaned, "\n")
+}