chore(sanitizer): remove explanatory reasoning comments

Keep the tool/function-call preservation logic intact while removing noisy implementation comments from the PR diff.

Co-Authored-By: OpenClaude (dmassoneto) <openclaude@gitlawb.com>

open-sse/handlers/responseSanitizer.ts

@@ -285,10 +285,6 @@ function sanitizeMessage(msg: unknown): unknown {
   // Non-streaming responses should not expose both visible content and reasoning_content.
   // Some clients drop the visible assistant text or render duplicated panels when both fields
   // are present in the final payload. Keep reasoning_content only for reasoning-only messages.
-  // EXCEPTION: When tool_calls (or legacy function_call) are present, reasoning_content
-  // must be preserved because thinking-enabled providers (e.g., Kimi) require it on
-  // assistant tool call messages. Without it, subsequent requests fail with:
-  //   "thinking is enabled but reasoning_content is missing in assistant tool call message"
   if (
     sanitized.reasoning_content !== undefined &&
     hasVisibleMessageContent(sanitized.content) &&

tests/unit/response-sanitizer.test.ts

@@ -58,8 +58,6 @@ test("sanitizeOpenAIResponse extracts thinking, collapses newlines, preserves re
   assert.equal((sanitized as any).choices[0].index, 2);
   assert.equal((sanitized as any).choices[0].finish_reason, "tool_calls");
   (assert as any).equal((sanitized as any).choices[0].message.content, "Hello\n\nworld");
-  // reasoning_content extracted from <think> tags is preserved when tool_calls exist
-  // because thinking-enabled providers require it on assistant tool call messages
   assert.equal((sanitized as any).choices[0].message.reasoning_content, "internal chain");
   (assert as any).deepEqual((sanitized as any).choices[0].message.tool_calls, [{ id: "call_1" }]);
   assert.deepEqual((sanitized as any).choices[0].message.function_call, { name: "legacy" });