Manual merge PR #1879: Fix indent inconsistency in ApiProxy templates and update docs, preserving v4.1.16

Author: lbjlaq

README.md

@@ -8,7 +8,7 @@
 
   <p>
     <a href="https://github.com/lbjlaq/Antigravity-Manager">
-      <img src="https://img.shields.io/badge/Version-4.1.16-blue?style=flat-square" alt="Version">
+      <img src="https://img.shields.io/badge/Version-4.1.15-blue?style=flat-square" alt="Version">
     </a>
     <img src="https://img.shields.io/badge/Tauri-v2-orange?style=flat-square" alt="Tauri">
     <img src="https://img.shields.io/badge/Backend-Rust-red?style=flat-square" alt="Rust">
@@ -413,10 +413,6 @@ response = client.chat.completions.create(
 ## 📝 开发者与社区
 
 *   **版本演进 (Changelog)**:
-    *   **v4.1.16 (2026-02-12)**:
-        -   **[核心修复] 解决 Gemini 图像生成因关键词匹配导致的 effortLevel 冲突 (PR #1873)**:
-            -   **逻辑冲突修复**: 彻底修复了 `gemini-3-pro-image` 及其 4k/2k 变体因包含 `gemini-3-pro` 关键词，被系统错误判定为支持 Adaptive Thinking 从而误注入 `effortLevel` 导致的 HTTP 400 错误。
-            -   **参数清洗**: 在代理请求层增加了对图像生成模型的特殊过滤，确保不再为非思维链模型注入不兼容的生成参数。
     *   **v4.1.15 (2026-02-11)**:
         -   **[核心功能] 开启 macOS 与 Windows 原生自动更新支持 (PR #1850)**:
             -   **端到端自动更新**: 启用了 Tauri 的原生更新插件，支持在应用内直接检测、下载并安装更新。

README_EN.md

@@ -138,7 +138,7 @@ curl -sSL https://raw.githubusercontent.com/lbjlaq/Antigravity-Manager/main/depl
 
 **Option 2: via Homebrew** (If you have [Linuxbrew](https://sh.brew.sh/) installed)
 ```bash
-brew tap lbjlaq/antigravity-manager https://github.com/lbjlaq/Antigravity-Manager/releases/download/v4.1.16/Antigravity_Tools_4.1.16_x64.dmg
+brew tap lbjlaq/antigravity-manager https://github.com/lbjlaq/Antigravity-Manager/releases/download/v4.1.15/Antigravity_Tools_4.1.15_x64.dmg
 ```
 
 #### Other Linux Distributions
@@ -264,10 +264,6 @@ print(response.choices[0].message.content)
 ## 📝 Developer & Community
 
 *   **Changelog**:
-    *   **v4.1.16 (2026-02-12)**:
-        -   **[Core Fix] Resolve effortLevel conflict in Gemini Image Generation caused by keyword matching (PR #1873)**:
-            -   **Logic Conflict Fix**: Completely fixed the HTTP 400 error where `gemini-3-pro-image` and its 4k/2k variants were incorrectly identified as supporting Adaptive Thinking due to the `gemini-3-pro` keyword, leading to the erroneous injection of `effortLevel`.
-            -   **Parameter Scrubbing**: Added specific filtering for image generation models at the proxy layer to ensure incompatible generation parameters are no longer injected into non-thinking models.
     *   **v4.1.15 (2026-02-11)**:
         -   **[Core Feature] Enable Native Auto-Update for macOS and Windows (PR #1850)**:
             -   **End-to-End Auto-Update**: Enabled the native Tauri updater plugin, supporting in-app update checks, downloads, and installations.

docker/README.md

@@ -126,7 +126,7 @@ docker build --build-arg USE_MIRROR=true -t antigravity-manager:latest -f docker
 ```bash
 # 打上版本標籤並推送
 docker tag antigravity-manager:latest lbjlaq/antigravity-manager:latest
-docker tag antigravity-manager:latest lbjlaq/antigravity-manager:4.1.16
+docker tag antigravity-manager:latest lbjlaq/antigravity-manager:4.1.15
 docker push lbjlaq/antigravity-manager:latest
-docker push lbjlaq/antigravity-manager:4.1.16
+docker push lbjlaq/antigravity-manager:4.1.15
 ```

src/pages/ApiProxy.tsx

@@ -959,21 +959,21 @@ export default function ApiProxy() {
         // 1. Anthropic Protocol
         if (selectedProtocol === 'anthropic') {
             return `from anthropic import Anthropic
- 
- client = Anthropic(
-     # 推荐使用 127.0.0.1
-     base_url="${`http://127.0.0.1:${port}`}",
-     api_key="${apiKey}"
- )
- 
- # 注意: Antigravity 支持使用 Anthropic SDK 调用任意模型
- response = client.messages.create(
-     model="${modelId}",
-     max_tokens=1024,
-     messages=[{"role": "user", "content": "Hello"}]
- )
- 
- print(response.content[0].text)`;
+
+client = Anthropic(
+    # 推荐使用 127.0.0.1
+    base_url="${`http://127.0.0.1:${port}`}",
+    api_key="${apiKey}"
+)
+
+# 注意: Antigravity 支持使用 Anthropic SDK 调用任意模型
+response = client.messages.create(
+    model="${modelId}",
+    max_tokens=1024,
+    messages=[{"role": "user", "content": "Hello"}]
+)
+
+print(response.content[0].text)`;
         }
 
         // 2. Gemini Protocol (Native)
@@ -997,43 +997,43 @@ print(response.text)`;
         // 3. OpenAI Protocol
         if (modelId.startsWith('gemini-3-pro-image')) {
             return `from openai import OpenAI
- 
- client = OpenAI(
-     base_url="${baseUrl}",
-     api_key="${apiKey}"
- )
- 
- response = client.chat.completions.create(
-     model="${modelId}",
-     # 方式 1: 使用 size 参数 (推荐)
-     # 支持: "1024x1024" (1:1), "1280x720" (16:9), "720x1280" (9:16), "1216x896" (4:3)
-     extra_body={ "size": "1024x1024" },
-     
-     # 方式 2: 使用模型后缀
-     # 例如: gemini-3-pro-image-16-9, gemini-3-pro-image-4-3
-     # model="gemini-3-pro-image-16-9",
-     messages=[{
-         "role": "user",
-         "content": "Draw a futuristic city"
-     }]
- )
- 
- print(response.choices[0].message.content)`;
+
+client = OpenAI(
+    base_url="${baseUrl}",
+    api_key="${apiKey}"
+)
+
+response = client.chat.completions.create(
+    model="${modelId}",
+    # 方式 1: 使用 size 参数 (推荐)
+    # 支持: "1024x1024" (1:1), "1280x720" (16:9), "720x1280" (9:16), "1216x896" (4:3)
+    extra_body={ "size": "1024x1024" },
+    
+    # 方式 2: 使用模型后缀
+    # 例如: gemini-3-pro-image-16-9, gemini-3-pro-image-4-3
+    # model="gemini-3-pro-image-16-9",
+    messages=[{
+        "role": "user",
+        "content": "Draw a futuristic city"
+    }]
+)
+
+print(response.choices[0].message.content)`;
         }
 
         return `from openai import OpenAI
- 
- client = OpenAI(
-     base_url="${baseUrl}",
-     api_key="${apiKey}"
- )
- 
- response = client.chat.completions.create(
-     model="${modelId}",
-     messages=[{"role": "user", "content": "Hello"}]
- )
- 
- print(response.choices[0].message.content)`;
+
+client = OpenAI(
+    base_url="${baseUrl}",
+    api_key="${apiKey}"
+)
+
+response = client.chat.completions.create(
+    model="${modelId}",
+    messages=[{"role": "user", "content": "Hello"}]
+)
+
+print(response.choices[0].message.content)`;
     };
 
     // 在 filter 逻辑中，当选择 openai 协议时，允许显示所有模型