diff --git a/bin/gitea-mcp b/bin/gitea-mcp
deleted file mode 100644
index 68f84c9..0000000
Binary files a/bin/gitea-mcp and /dev/null differ
diff --git a/worker.js b/worker.js
index 7a30100..9a265dd 100644
--- a/worker.js
+++ b/worker.js
@@ -3,17 +3,17 @@ export default {
     const url = new URL(request.url);
     const { pathname } = url;
 
-    // CORS preflight for API
+    // API 的 CORS 预检请求
     if (request.method === "OPTIONS" && pathname.startsWith("/api/")) {
       return new Response(null, { headers: corsHeaders() });
     }
 
-    // --- API: list languages (common) ---
+    // --- API：返回语言列表（常用） ---
     if (pathname === "/api/languages") {
       return jsonResponse({ languages: COMMON_LANGUAGES }, 200, corsHeaders());
     }
 
-    // --- API: translate ---
+    // --- API：翻译 ---
     if (pathname === "/api/translate") {
       if (request.method !== "POST") {
         return jsonResponse(
@@ -23,7 +23,7 @@ export default {
         );
       }
 
-      // Optional API key
+      // 可选的 API Key 校验
       const apiKey = env.API_KEY;
       if (apiKey) {
         const got =
@@ -44,7 +44,7 @@ export default {
 
       const text = (body.text ?? "").toString();
       const target_lang = (body.target_lang ?? "").toString().trim();
-      const source_lang_raw = (body.source_lang ?? "").toString().trim(); // optional
+      const source_lang_raw = (body.source_lang ?? "").toString().trim(); // 可选
 
       if (!text) return jsonResponse({ error: "text is required" }, 400, corsHeaders());
       if (!target_lang) return jsonResponse({ error: "target_lang is required" }, 400, corsHeaders());
@@ -59,7 +59,7 @@ export default {
 
       try {
         if (translate_all) {
-          // Use batch requests (Workers AI supports batch schema on this model) :contentReference[oaicite:2]{index=2}
+          // 使用批量请求（Workers AI 在该模型上支持批量 schema）
           const requests = chunks.slice(0, 50).map((t) => ({
             text: t,
             ...(source_lang_raw ? { source_lang: normalizeLang(source_lang_raw) } : {}),
@@ -67,12 +67,12 @@ export default {
           }));
 
           const out = await env.AI.run("@cf/meta/m2m100-1.2b", { requests });
-          // Some bindings return array-like results; normalize defensively:
+          // 某些绑定会返回类似数组的结果；这里做防御性归一化处理：
           const translated_pages = Array.isArray(out)
             ? out.map((x) => x?.translated_text ?? "")
             : (out?.responses ?? out?.results ?? out?.translated_pages ?? null);
 
-          // Fallback: if unknown, just translate one-by-one
+          // 兜底处理：如果返回结构不符合预期，就退回到单页翻译
           if (Array.isArray(translated_pages)) {
             return jsonResponse(
               {
@@ -89,7 +89,7 @@ export default {
           }
         }
 
-        // Default: translate one page
+        // 默认：仅翻译单页
         const chunk = chunks[safePage - 1] ?? "";
         const out = await env.AI.run("@cf/meta/m2m100-1.2b", {
           text: chunk,
@@ -119,7 +119,7 @@ export default {
       }
     }
 
-    // --- Web UI ---
+    // --- Web 界面 ---
     if (pathname === "/" || pathname === "/index.html") {
       if (request.method === "GET") {
         return new Response(renderPage({}), {
@@ -182,7 +182,7 @@ export default {
   },
 };
 
-// ---------------- helpers ----------------
+// ---------------- 辅助函数 ----------------
 
 const COMMON_LANGUAGES = [
   { code: "en", name: "English" },
@@ -208,8 +208,8 @@ const COMMON_LANGUAGES = [
 ];
 
 function normalizeLang(lang) {
-  // Docs say language code like 'en'/'es' etc. :contentReference[oaicite:3]{index=3}
-  // But examples sometimes use english/french; we accept both by passing through.
+  // 文档里写的是语言代码，例如 'en'/'es' 等。
+  // 但有些示例会使用 english/french，这里也允许原样透传。
   return String(lang || "").trim().toLowerCase();
 }
 
@@ -223,7 +223,7 @@ function splitText(text, maxChars) {
   const t = String(text || "");
   if (t.length <= maxChars) return [t];
 
-  // Prefer paragraph-based splitting
+  // 优先按段落拆分
   const parts = t.split(/\n{2,}/);
   const chunks = [];
   let buf = "";
@@ -231,7 +231,7 @@ function splitText(text, maxChars) {
   for (const p of parts) {
     const piece = (p + "\n\n");
     if (piece.length > maxChars) {
-      // Hard split long paragraph
+      // 对超长段落进行硬切分
       if (buf) {
         chunks.push(buf);
         buf = "";