GenarrativeAI/Genarrative
11
0
Fork 0
Code Issues 21 Pull Requests 4 Actions Packages Projects 3 Releases Wiki Activity
2,618 Commits 24 Branches 4 Tags
84ec79a8bbf96fc41aa188214ffbe80a2fceb310
Commit Graph

40 Commits

Author SHA1 Message Date
kdletters 6f181b36a1 修复编辑器生成幂等与参考图契约
Project CI / Frontend tests (push) Failing after 22s
Details
Project CI / Repository checks (push) Successful in 58s
Details
Project CI / Backend tests (push) Successful in 3m30s
Details
Project CI / Native shell tests (push) Successful in 12m34s
Details 
生成队列按稳定请求标识去重并保持外部幂等哈希兼容
统一拒绝参考图超限并同步前端、后端、Provider 与 OpenAPI 契约
按真实归属重建生成引用并阻止直接持久化伪造来源
锁定参考图在途上传上下文并保留批量部分成功结果
关闭内部生成 POST 自动重试并补齐回归测试与项目文档
修正最新主线开发者密钥弹窗的导入排序门禁
 2026-08-05 23:31:16 +08:00
lhk229 8f19964e3f 加入手动完美像素入口；像素艺术勾选时自动补提示词约束 (#124)
Project CI / Repository checks (push) Successful in 1m0s
Details
Project CI / Backend tests (push) Successful in 3m52s
Details
Project CI / Native shell tests (push) Successful in 13m8s
Details
Project CI / Frontend tests (push) Successful in 2m43s
Details 
Co-authored-by: 段舒康 <kdletters@qq.com>
Reviewed-on: http://192.168.35.82/git/GenarrativeAI/Genarrative/pulls/124
Reviewed-by: 段舒康 <kdletters@qq.com>
Co-authored-by: Linghong <ink29535@proton.me>
Co-committed-by: Linghong <ink29535@proton.me>
 2026-08-05 21:28:40 +08:00
kdletters 842d99697a 合并游戏聊天0.1.1到master
Project CI / Repository checks (push) Successful in 1m11s
Details
Project CI / Frontend tests (push) Failing after 2m6s
Details
Project CI / Backend tests (push) Failing after 2m44s
Details
Project CI / Native shell tests (push) Failing after 3m14s
Details 
合入独立game-chat release构建、内置Runner与退出收束。
合入自动预览刷新、结构化进度展示及强化试玩验收。
保留master资源画布、共享预览和原生壳严格门禁。
修复Windows Runner单实例锁与GUI owner锁固定路径兼容。
 2026-07-31 14:27:36 +08:00
kdletters cf1a023125 收紧图集切片内存并批量持久化
Project CI / Repository checks (push) Successful in 1m0s
Details
Project CI / Backend tests (push) Failing after 2m42s
Details
Project CI / Frontend tests (push) Successful in 2m45s
Details
Project CI / Native shell tests (push) Successful in 11m42s
Details 
按需编码并限制图集上传并发、超时与累计裁剪像素
批量确认切片对象、项目资源、账号素材和完成批次
稳定派生切片记录标识并收紧重放与来源资源校验
复用已鉴权项目资源避免手动拆分全账号扫描
补充图集资源边界文档与回归测试
 2026-07-31 13:40:28 +08:00
menghao 216407d93e 实现资源画布布局持久化 (#116)
Project CI / Frontend tests (push) Failing after 20s
Details
Project CI / Repository checks (push) Successful in 1m2s
Details
Project CI / Backend tests (push) Successful in 2m59s
Details
Project CI / Native shell tests (push) Successful in 11m58s
Details 
冻结资源画布布局数据与 CAS 合同
实现双模式本地 sidecar 安全读写
接入二维拖动、默认排版和跨重启恢复
补齐并发冲突、安全边界和界面测试
同步技术文档与共享决策

Reviewed-on: http://192.168.35.82/git/GenarrativeAI/Genarrative/pulls/116
Reviewed-by: 段舒康 <kdletters@qq.com>
Co-authored-by: menghao <mh18530625731@163.com>
Co-committed-by: menghao <mh18530625731@163.com>
 2026-07-31 12:00:48 +08:00
kdletters b512430e53 收紧图集切片资源边界
Project CI / Repository checks (push) Successful in 1m4s
Details
Project CI / Frontend tests (push) Successful in 2m5s
Details
Project CI / Native shell tests (push) Successful in 2m32s
Details
Project CI / Backend tests (push) Successful in 3m37s
Details 
前置限制原始连通域、局部候选和输出切片数量
统一图集切片的有界阻塞线程执行
统一角色图标和UI的原图安全降级
补齐异常输入回归测试与工程文档
 2026-07-30 13:44:17 +08:00
lhk229 8848a68fb0 调整像素网格步长估算分位数 
将相邻边缘峰间距由中位数改为线性插值 P30。

同步更新像素规整技术方案和共享决策记录。
 2026-07-29 05:42:08 +00:00
lhk229 1ed8064d2f 统一图标图集全连通域拆分 
移除前端从提示词解析素材数量的逻辑，完整提示词作为单个请求元素提交。

删除后端按描述数量截断切片的路径，自动拆分与手动拆分复用全连通域算法。

统一切片命名和尺寸、像素、数量限制，保留手动拆分按钮与接口。

更新既有测试、OpenAPI、编辑器文档和共享决策记录。
 2026-07-29 04:52:29 +00:00
lhk229 732fcf0b45 新增图片生成像素艺术后处理 
新增确定性像素网格规整模块并保留第三方 MIT 许可证。

为普通图片、角色和图标图集接入像素艺术选项、并发预算和失败降级。

同步前端交互、请求契约、OpenAPI、项目文档与测试。
 2026-07-29 03:46:03 +00:00
k88936 66a8b7f597 画布agent 完善上下文 (#101)
Project CI / Repository checks (push) Successful in 53s
Details
Project CI / Frontend tests (push) Successful in 2m34s
Details
Project CI / Native shell tests (push) Successful in 2m42s
Details
Project CI / Backend tests (push) Successful in 3m28s
Details 
* 补全完整的工具参数
* 调整提示词应对一些bad case
* 把attachment的名称/描述纳入上下文

Reviewed-on: http://192.168.35.82/git/GenarrativeAI/Genarrative/pulls/101
Co-authored-by: 王德宇 <kvtodev@outlook.com>
Co-committed-by: 王德宇 <kvtodev@outlook.com>
 2026-07-28 15:44:05 +08:00
kdletters 31a93c6842 补齐Gitea测试门禁
Project CI / Repository checks (push) Successful in 2m9s
Details
Project CI / Frontend tests (push) Successful in 1m22s
Details
Project CI / Backend tests (push) Successful in 7m25s
Details
Project CI / Native shell tests (push) Successful in 5m56s
Details 
拆分前端后端与原生壳测试任务
接入server-rs正式workspace全量测试
修正过期Rust断言和Vitest遗漏路径
将AI游戏创作分支纳入CI触发范围
 2026-07-21 18:12:43 +08:00
kdletters 6bed37542c 调整 VectorEngine 图片模型兜底顺序 
统一 gpt-image-2 首发并在符合条件的失败后切换 gpt-image-2-c
保留总尝试预算、worker deadline 与首选失败审计
补齐响应图片校验、脚本兜底和回归测试
同步后端架构、开发运维和项目共享记忆文档
 2026-07-21 17:06:39 +08:00
kdletters f748dc72c3 收口图片生成任务超时预算 
- 将四类 VectorEngine 图片任务纳入长任务预算并预留终态写回窗口
- 通过请求上下文传递绝对截止时间并约束发送重试与图片下载
- 允许显式下调单次请求超时并补齐预算回归测试
- 同步外部生成 Worker 运维文档与项目记忆
 2026-07-20 20:57:34 +08:00
menghao 47f3297e84 修复图片改造后二维清晰度降级问题 (#90) 
前端按图层真实尺寸恢复图片改造比例与清晰度

后端在角色去背景后恢复生成原图像素尺寸

补充前后端二千像素尺寸回归测试

同步更新图片画布技术文档与项目决策记录

---------

Co-authored-by: MengHao <menghao@MengHaodeMacBook-Air.local>
Reviewed-on: https://git.genarrative.world/git/GenarrativeAI/Genarrative/pulls/90
Reviewed-by: 段舒康 <kdletters@qq.com>
Co-authored-by: 孟豪 <mh18530625731@163.com>
Co-committed-by: 孟豪 <mh18530625731@163.com>
 2026-07-18 16:50:17 +08:00
k88936 34c6ed8756 自动扩缩edit图片的大小适应gptimage2的生成图片大小参数要求 (#82) 
尺寸严格限制规则 (Strict Size Constraints):
图片最大边长 ≤ 3840px <- by hardcoded length
宽高两边像素均为 16px 的倍数 <- by backend do crop
长边 / 短边 比值 ≤ 3:1 <- by hardcoded aspect ratio
总像素范围：最小 655360 ~ 最大 8294400 <- by auto scale in this pr

logic impl in the backend

---------

Co-authored-by: 段舒康 <kdletters@qq.com>
Reviewed-on: https://git.genarrative.world/git/GenarrativeAI/Genarrative/pulls/82
Co-authored-by: kvtodev <kvtodev@outlook.com>
Co-committed-by: kvtodev <kvtodev@outlook.com>
 2026-07-14 21:04:54 +08:00
kdletters eae00292a1 新增画板图集自动拆分 
生成图标图集后自动拆分并持久化独立素材
新增图集工具栏手动拆分入口与权威画布回填
修复连通域切片阅读顺序并补齐前后端测试
同步外部接口与编辑器设计文档
 2026-07-13 19:10:30 +08:00
kdletters 53c5639620 完善画布任务与标准绿幕抠除链路 
统一编辑器标准绿幕提示词和本地绿幕透明化后处理

生成标准绿幕资产前保存源图到 OSS 并下沉字节级抠绿到 platform-image

将手动去背景接入外部生成队列与画布任务侧栏

修复任务侧栏分页、滚动、计时、焦点与完成信息展示

补充右键菜单关闭和素材复制下载交互

同步编辑器相关文档和项目记忆

(cherry picked from commit a8c19e132d)
 2026-06-30 17:12:13 +08:00
kdletters d73dc02f4d 合并 master 到跳一跳分支 
合入 origin/master 最新变更

解决跳一跳提示词测试与图集去背冲突
 2026-06-26 21:30:26 +08:00
kdletters 4b2cac44ff 修复图标素材切片废图 
合并靠近主体的断开高光阴影连通域

过滤孤立小碎片避免生成废图素材

补充图片画布图标切片踩坑记录
 2026-06-23 19:30:04 +08:00
kdletters c206ca7c33 固定绿幕提示词颜色 
统一绿幕生成提示词使用 #00FF00 / RGB(0,255,0)

补充暗绿绿幕去背兜底并覆盖聚焦测试

同步玩法、后端与项目记忆文档口径
 2026-06-22 15:11:33 +08:00
JenkenB 160c96a014 合并画板素材生成与宣发入口改动 
新增 UI 设计图提取素材入口并复用 gpt-image-2 spritesheet 拆分流程。

保留图标生成与 UI 提取的 spritesheet 原图并同步放置拆分素材到画布。

补齐宣发素材演示入口、占位图样式、引用上传与生成提交链路。

调整编辑器生成接口 body 限制和 VectorEngine 请求兼容逻辑。

更新画板音乐生成、项目基线、宣发素材与图片画布相关文档。
 2026-06-19 10:37:22 +08:00
JenkenB 097170836d Add UI-design asset extraction & spritesheet support 
Introduce a new UI design asset extraction flow: add POST /api/editor/ui-designs/assets/extractions handler that always uses gpt-image-2 with the fixed prompt "提取画面中的所有独立并整理成spritesheet"; parse Data URL references, call OpenAI image edits, return a spritesheet and sliced icon assets. Increase request body limit for image Data URLs and wire the route with bearer auth. Add platform-image slicing: slice_generated_icon_spritesheet_all_by_connected_components (with auto-naming "素材 1..N", foreground estimation and heuristics) and export it. Add Rust tests for the handler, prompt constant, and slicing behavior. Update frontend client and integration tests to call the new extractEditorUiDesignAssets API and verify spritesheet + icon layers are added to the canvas and persisted. Update multiple docs to describe the extraction UI, toolbar placement, and related audio/playback pitfalls. Minor related updates: toolbar button ordering and audio duration / signed read-url notes.
 2026-06-19 10:09:19 +08:00
kdletters ae375a8e5f 合并编辑器素材库能力 
合并 codex/editor-asset-library 到宣发素材演示分支

保留素材库、生成器、音频、UI设计图和持久化相关能力

补回宣发素材入口与 5 泥点计费契约
 2026-06-18 22:36:54 +08:00
JenkenB 9c7b3ad16d Image editor: Seedance2.0 refs and audio changes 
Adds Seedance 2.0 reference-media constraints and forbids video data URLs; updates docs/decision-log and pitfalls to record the new rules. Introduces typed audio params (type: one-shot|loop, tempo: number|null), generator-specific placeholder styles/avoidance and UI changes for generation panels. Implements related front-end changes (new crop/ raster edit components, many image-editor models, tests and workflow hooks) and server-side contract/handler updates across api-server, shared-contracts, platform-audio, platform-image and vector_engine_audio_generation to support the new submission/asset flow and validations.
 2026-06-18 17:31:35 +08:00
Codex 638dc9f6f6 收口图片编辑器宣发素材生成 
调整详情五图工作流为单张竖版详情图生成

修正宣发素材明确像素尺寸透传和运营海报尺寸映射

压缩宣发参考图提交并提高后端请求体限制

补充前后端回归测试、文档和 Hermes 踩坑记录
 2026-06-18 15:49:20 +08:00
JenkenB b2fd5574db 接入画板生成视频功能 
新增画板底部生成视频入口、Lovart 风格面板、视频图层渲染与元数据展示。

接入 /api/editor/videos/generations 契约与后端 Ark/VectorEngine 视频任务链路。

统一编辑器生成类泥点配置，并补充 UI 设计图、参考图与生成面板结构测试。

更新编辑器技术方案、生成类面板方案和 Hermes 共享决策/踩坑记录。
 2026-06-17 20:52:45 +08:00
JenkenB 7eeff10c67 新增编辑器生成规范、生成角色形象、生成图标素材等功能 
新增编辑器生成规范、生成角色形象、生成图标素材等功能
 2026-06-16 14:47:13 +08:00
lhk229 95df62fc82 跳一跳UV面切分改用blob+gradient自适应算法 
重构alpha.rs洋红去背预设参数
新增jump_hop_atlas_slicing.rs独立切图模块
修复jump_hop.rs调用链接入新切图算法
 2026-06-10 19:49:39 +08:00
wuxiangwanzi a0473771f1 fix: 优化跳一跳运行态与地块资源 2026-06-09 01:28:30 +08:00
lhk229 1dcba515b2 add internal holes detection in jump-hop image processing 2026-06-08 21:32:44 +08:00
wuxiangwanzi a215852381 Merge remote-tracking branch 'origin/master' into codex/tiaoyitiao 2026-06-05 22:46:57 +08:00
kdletters ed6a59e641 fix: send VectorEngine edit images as file parts 2026-06-05 21:19:46 +08:00
kdletters 5150925947 fix: send VectorEngine images with libcurl 2026-06-05 17:03:49 +08:00
wuxiangwanzi 0edcb1b9f1 fix(jump-hop): clean magenta cutout fringes 2026-06-05 13:23:58 +08:00
kdletters 9ab353926e fix: retry VectorEngine image send timeouts 2026-06-05 11:38:02 +08:00
wuxiangwanzi 0041b95f72 feat(jump-hop): optimize generated assets and runtime background 2026-06-04 22:34:19 +08:00
kdletters 2678954627 fix: log VectorEngine image edit request params 2026-06-04 01:22:28 +08:00
kdletters 08577b66c5 fix: stabilize puzzle vector engine asset generation 2026-06-03 02:40:07 +08:00
kdletters 44c65df5c9 refactor: extract platform media crates 2026-05-26 13:18:13 +08:00
kdletters 080694fb46 refactor: extract platform image provider 2026-05-25 19:03:43 +08:00