Switch to VectorEngine gpt-image-2 and edits

Replace uses of the legacy `gpt-image-2-all` model with `gpt-image-2` and standardize image workflows: no-reference generation uses POST /v1/images/generations, any-reference flows use POST /v1/images/edits with multipart `image` parts. Update SKILLs, generation scripts, decision logs, and docs to reflect the contract change and edits-vs-generations guidance. Apply corresponding changes across backend (api-server match3d/puzzle modules, openai image adapter, mappers, telemetry, spacetime client/module), frontend components and services (Match3D, Puzzle, CreativeImageInputPanel, runtime shells), and add new spritesheet/parser files and tests. Also add media/logo.png. These changes align repository code and documentation with the VectorEngine image API contract and update generation/upload handling (green-screen -> alpha processing, spritesheet handling, and related tests).

server-rs/crates/api-server/src/puzzle.rs

@@ -1,6 +1,5 @@
 use std::{
     collections::BTreeMap,
-    error::Error as StdError,
     time::{Duration, Instant, SystemTime, UNIX_EPOCH},
 };
 
@@ -78,10 +77,11 @@ use crate::{
         should_skip_asset_operation_billing_for_connectivity,
     },
     auth::AuthenticatedAccessToken,
+    generated_asset_sheets::apply_generated_asset_sheet_green_screen_alpha,
     http_error::AppError,
     llm_model_routing::{CREATION_TEMPLATE_LLM_MODEL, PUZZLE_LEVEL_NAME_VISION_LLM_MODEL},
     openai_image_generation::{
-        DownloadedOpenAiImage, VECTOR_ENGINE_GPT_IMAGE_2_MODEL, build_openai_image_http_client,
+        DownloadedOpenAiImage, GPT_IMAGE_2_MODEL, build_openai_image_http_client,
         create_openai_image_generation, require_openai_image_settings,
     },
     platform_errors::map_oss_error,
@@ -124,9 +124,13 @@ const PUZZLE_LEVEL_NAME_VISION_IMAGE_MAX_SIDE: u32 = 768;
 const PUZZLE_LEVEL_NAME_VISION_MAX_TOKENS: u32 = 512;
 const PUZZLE_REFERENCE_IMAGE_MAX_BYTES: usize = 8 * 1024 * 1024;
 const PUZZLE_REFERENCE_IMAGE_SOURCE_LIMIT: usize = 5;
-const PUZZLE_VECTOR_ENGINE_IMAGE_EDIT_MODEL: &str = "gpt-image-2";
 const PUZZLE_UI_BACKGROUND_PROMPT_FALLBACK_MARKER: &str =
     "移动端拼图游戏纯背景，题材氛围清晰，不包含拼图槽或 UI 元素";
+const PUZZLE_VECTOR_ENGINE_SQUARE_IMAGE_SIZE: &str = "1024x1024";
+const PUZZLE_VECTOR_ENGINE_PORTRAIT_IMAGE_SIZE: &str = "1024x1536";
+const PUZZLE_LEVEL_SCENE_IMAGE_PROMPT: &str = "参考图作为拼图画面，生成对应的拼图游戏关卡画面，要求画面中所有元素精致且风格高度一致，画面中所有UI细节饱满精致、完成度高、顶级游戏品质\n\n画面元素：\n返回按钮位于顶部左上角，顶部中间显示关卡标题“第1关 影”和倒计时时间，右上角显示设置按钮\n画面中间是一个正方形的3*3拼图，拼图区域宽度与画面宽度同宽，紧贴画面横向边缘，拼图区域边界带有边框装饰\n拼图区域下方包含一个下一关按钮，仅在关卡完成时显示\n底部是三个贴合画面主题的道具按钮分别为“提示”、“原图”、“冻结”\n道具按钮上不要显示次数标注，返回按钮和设置按钮旁禁止标注文字";
+const PUZZLE_UI_SPRITESHEET_IMAGE_PROMPT: &str = "提取画面中的UI元素，将返回按钮、设置按钮、下一关按钮、提示按钮、原图按钮、冻结按钮整理成纯绿色绿幕背景的spritesheet。背景必须是统一纯绿色绿幕（高饱和亮绿色，接近 #00FF00），背景平整无纹理、无渐变、无阴影、无场景内容，后端会在生图后将绿幕扣成透明并把透明背景 PNG 存到 OSS。按钮顺序必须按原图位置从左到右、从上到下排列：返回、设置、下一关、提示、原图、冻结。按钮素材内必须保留对应中文文字，每个按钮必须是独立完整图形，按钮之间保留足够纯绿色绿幕空白，不能相互接触、重叠或连成一片，方便运行态按自动边界检测识别矩形素材。返回按钮和设置按钮不要额外画白色外圈、白底圆环或浮雕外框，直接画扁平图标本体。按钮自身不得使用接近 #00FF00 的高饱和纯绿；绿色题材只能使用深绿、橄榄绿、金绿或蓝绿，并用清晰描边与绿幕区分。禁止水印、数字、次数标注、透明背景、背景图、拼图块、棋盘、网格线、按钮外标签和额外按钮。";
+const PUZZLE_LEVEL_BACKGROUND_IMAGE_PROMPT: &str = "移除参考图中所有UI元素、移除拼图画面，仅保留背景图，补全被覆盖的背景图内容。禁止在背景中出现人像或和拼图画面中主体一致的内容";
 mod handlers;
 pub(crate) use self::handlers::*;