feat: support OpenAI tools call

.github/ISSUE_TEMPLATE/tool_call_error_cn.yml

@@ -0,0 +1,67 @@
+name: 工具调用错误报告
+description: 报告工具调用相关的错误 (中文)
+title: "[工具错误]: "
+labels: ["🐛 Bug"]
+body:
+  - type: markdown
+    attributes:
+      value: |
+        如果您遇到了工具调用（Tool Calls）相关的错误，请务必按照以下格式提交一个新的 Issue，以便我们能够快速定位和解决问题。
+
+        **请注意：** 为了排查工具调用问题，我们需要您提供详细的上下文日志。请在复现问题时开启调试模式，并提供完整的请求和响应日志。
+
+  - type: input
+    id: version
+    attributes:
+      label: 版本号
+      description: 您当前使用的版本
+      placeholder: "例如: v0.2.11"
+    validations:
+      required: true
+
+  - type: dropdown
+    id: api_endpoint
+    attributes:
+      label: 请求接口
+      options:
+        - OpenAI
+        - Gemini
+    validations:
+      required: true
+
+  - type: input
+    id: model_name
+    attributes:
+      label: 模型名称
+      placeholder: "例如: gemini-3-pro-preview / gemini-3-flash-preview"
+    validations:
+      required: true
+
+  - type: input
+    id: client_tool
+    attributes:
+      label: 客户端/工具
+      placeholder: "例如: NextChat / LobeChat / OpenCode"
+    validations:
+      required: true
+
+  - type: textarea
+    id: problem_description
+    attributes:
+      label: 问题描述
+      description: "请简要描述您遇到的错误现象，例如：请求成功但参数解析错误、返回 400 错误等"
+      placeholder: 描述错误的具体表现...
+    validations:
+      required: true
+
+  - type: textarea
+    id: logs
+    attributes:
+      label: 日志信息 (关键)
+      description: |
+        **请务必开启调试模式 (Debug Mode) 并重新发送请求以获取完整日志。** 请提供**该次请求的所有相关日志**（可以通过日期/时间来确定范围）。
+        > ⚠️ **注意**: 请去除日志中的敏感信息（如 API Key），但**请务必保留日志的整体结构和格式**（尤其是 JSON 结构、空格和换行），这对排查解析错误至关重要。敏感内容可以用 `******` 替代。
+      render: text
+      placeholder: 在这里粘贴详细日志...
+    validations:
+      required: true

.github/ISSUE_TEMPLATE/tool_call_error_en.yml

@@ -0,0 +1,67 @@
+name: Tool Call Error Report
+description: Report an error related to tool calls (English)
+title: "[Tool Error]: "
+labels: ["🐛 Bug"]
+body:
+  - type: markdown
+    attributes:
+      value: |
+        If you encounter errors related to Tool Calls, please submit a new issue using the format below so that we can quickly identify and resolve the problem.
+
+        **Note:** To troubleshoot tool call issues, we need detailed context logs. Please enable debug mode when reproducing the issue and provide full request and response logs.
+
+  - type: input
+    id: version
+    attributes:
+      label: Version
+      description: The version you are currently using
+      placeholder: "e.g., v0.2.11"
+    validations:
+      required: true
+
+  - type: dropdown
+    id: api_endpoint
+    attributes:
+      label: API Endpoint
+      options:
+        - OpenAI
+        - Gemini
+    validations:
+      required: true
+
+  - type: input
+    id: model_name
+    attributes:
+      label: Model Name
+      placeholder: "e.g., gemini-3-pro-preview / gemini-3-flash-preview"
+    validations:
+      required: true
+
+  - type: input
+    id: client_tool
+    attributes:
+      label: Client/Tool
+      placeholder: "e.g., NextChat / LobeChat / OpenCode"
+    validations:
+      required: true
+
+  - type: textarea
+    id: problem_description
+    attributes:
+      label: Problem Description
+      description: "Briefly describe the error you encountered (e.g., request succeeded but parameter parsing failed, returned 400 error, etc.)"
+      placeholder: Describe the error details...
+    validations:
+      required: true
+
+  - type: textarea
+    id: logs
+    attributes:
+      label: Log Information (Critical)
+      description: |
+        **Please make sure to enable Debug Mode and resend the request to capture full logs.** Please provide **all logs related to this specific request** (you can determine the range by date/time).
+        > ⚠️ **Attention**: Please remove sensitive information (such as API Keys) from the logs, but **please preserve the overall structure and format of the logs** (especially JSON structure, spacing, and newlines). This is crucial for debugging parsing errors. You can replace sensitive content with `******`.
+      render: text
+      placeholder: Paste detailed logs here...
+    validations:
+      required: true

README.md

@@ -189,9 +189,10 @@ sudo docker compose down
 | `HOST`                      | 服务器监听的主机地址。                                        | `0.0.0.0`            |
 | `ICON_URL`                  | 用于自定义控制台的 favicon 图标。支持 ICO, PNG, SVG 等格式。  | `/AIStudio_logo.svg` |
 | `SECURE_COOKIES`            | 是否启用安全 Cookie。`true` 表示仅支持 HTTPS 协议访问控制台。 | `false`              |
-| `RATE_LIMIT_MAX_ATTEMPTS`   | 时间窗口内控制台允许的最大失败登录尝试次数（设为 0 禁用）。   | `5`                  |
+| `RATE_LIMIT_MAX_ATTEMPTS`   | 时间窗口内控制台允许的最大失败登录尝试次数（设为 `0` 禁用）。 | `5`                  |
 | `RATE_LIMIT_WINDOW_MINUTES` | 速率限制的时间窗口长度（分钟）。                              | `15`                 |
 | `CHECK_UPDATE`              | 是否在页面加载时检查版本更新。设为 `false` 可禁用。           | `true`               |
+| `LOG_LEVEL`                 | 日志输出等级。设为 `DEBUG` 启用详细调试日志。                 | `INFO`               |
 
 #### 🌐 代理配置
 
@@ -200,8 +201,8 @@ sudo docker compose down
 | `INITIAL_AUTH_INDEX`            | 启动时使用的初始身份验证索引。                       | `0`       |
 | `MAX_RETRIES`                   | 请求失败后的最大重试次数（仅对假流式和非流式生效）。 | `3`       |
 | `RETRY_DELAY`                   | 两次重试之间的间隔（毫秒）。                         | `2000`    |
-| `SWITCH_ON_USES`                | 自动切换帐户前允许的请求次数（设为 0 禁用）。        | `40`      |
-| `FAILURE_THRESHOLD`             | 切换帐户前允许的连续失败次数（设为 0 禁用）。        | `3`       |
+| `SWITCH_ON_USES`                | 自动切换帐户前允许的请求次数（设为 `0` 禁用）。      | `40`      |
+| `FAILURE_THRESHOLD`             | 切换帐户前允许的连续失败次数（设为 `0` 禁用）。      | `3`       |
 | `IMMEDIATE_SWITCH_STATUS_CODES` | 触发立即切换帐户的 HTTP 状态码（逗号分隔）。         | `429,503` |
 
 #### 🗒️ 其他配置

README_EN.md

@@ -189,9 +189,10 @@ This endpoint is forwarded to the official Gemini API format endpoint.
 | `HOST`                      | Server listening host address.                                                 | `0.0.0.0`            |
 | `ICON_URL`                  | Custom favicon URL for the console. Supports ICO, PNG, SVG, etc.               | `/AIStudio_logo.svg` |
 | `SECURE_COOKIES`            | Enable secure cookies. `true` for HTTPS only, `false` for both HTTP and HTTPS. | `false`              |
-| `RATE_LIMIT_MAX_ATTEMPTS`   | Maximum failed login attempts allowed within the time window (0 to disable).   | `5`                  |
+| `RATE_LIMIT_MAX_ATTEMPTS`   | Maximum failed login attempts allowed within the time window (`0` to disable). | `5`                  |
 | `RATE_LIMIT_WINDOW_MINUTES` | Time window for rate limiting in minutes.                                      | `15`                 |
 | `CHECK_UPDATE`              | Enable version update check on page load. Set to `false` to disable.           | `true`               |
+| `LOG_LEVEL`                 | Logging output level. Set to `DEBUG` for detailed debug logs.                  | `INFO`               |
 
 #### 🌐 Proxy Configuration
 
@@ -200,8 +201,8 @@ This endpoint is forwarded to the official Gemini API format endpoint.
 | `INITIAL_AUTH_INDEX`            | Initial authentication index to use on startup.                                                      | `0`       |
 | `MAX_RETRIES`                   | Maximum number of retries for failed requests (only effective for fake streaming and non-streaming). | `3`       |
 | `RETRY_DELAY`                   | Delay between retries in milliseconds.                                                               | `2000`    |
-| `SWITCH_ON_USES`                | Number of requests before automatically switching accounts (0 to disable).                           | `40`      |
-| `FAILURE_THRESHOLD`             | Number of consecutive failures before switching accounts (0 to disable).                             | `3`       |
+| `SWITCH_ON_USES`                | Number of requests before automatically switching accounts (`0` to disable).                         | `40`      |
+| `FAILURE_THRESHOLD`             | Number of consecutive failures before switching accounts (`0` to disable).                           | `3`       |
 | `IMMEDIATE_SWITCH_STATUS_CODES` | HTTP status codes that trigger immediate account switching (comma-separated).                        | `429,503` |
 
 #### 🗒️ Other Configuration