Skip to content

Releases: liwg1995/YuduBid

v0.9.6

Choose a tag to compare

@github-actions github-actions released this 26 Aug 06:33

中文说明

投标机会性能优化

  • 优化主动扫描期间的事件推送与界面刷新,降低高频状态更新造成的卡顿。
  • 批量扫描完成后仅刷新一次完整机会快照,避免重复查询、传输和整页渲染。
  • 减少公告入库过程中的重复正文读取、项目关联计算和监控匹配查询。
  • 优化机会快照统计与关联计数缓存,提升大规模本地机会库的加载和筛选速度。
  • 机会详情列表首批仅渲染 100 条,支持按需继续加载,使近千条机会下的滚动和交互更加流畅。

企业画像匹配

  • 扫描机会会根据企业画像中的目标行业、服务区域、核心能力、企业资质、人员证书、类似业绩和竞争优势生成本地匹配分。
  • 与企业画像匹配的项目自动优先显示,未匹配项目保持原有时间顺序并排列在后。
  • 总览和机会列表新增画像匹配分及匹配依据提示,修改企业画像后会立即重新计算排序。
  • 画像匹配完全在本地执行,不会额外调用 AI 模型或产生模型费用。

界面优化

  • 优化暗夜模式下画像匹配徽标的颜色、描边和层次,使高、中、低匹配状态更易辨识。
  • 优化暗夜模式下投标推进进度条的轨道和已完成部分,避免空进度显示为突兀的纯白色。

English Notes

Bid Opportunity Performance

  • Reduced UI stutter during active scans by optimizing progress events and renderer refreshes.
  • Refreshes the full opportunity snapshot only once after a batch scan completes, avoiding repeated queries, transfers, and full-page renders.
  • Removed redundant document reads, project-relation calculations, and monitor-match lookups during opportunity ingestion.
  • Improved snapshot aggregation and relation-count caching for faster loading and filtering in large local opportunity libraries.
  • Renders only the first 100 opportunity cards initially, with incremental loading for smoother scrolling and interaction across large datasets.

Enterprise Profile Matching

  • Generates a local match score from target industries, service regions, capabilities, qualifications, personnel certifications, past performance, and competitive advantages.
  • Automatically ranks profile-matched opportunities ahead of unmatched items while preserving the existing time order within the remaining results.
  • Shows profile match scores and evidence in both the overview and opportunity list, and recalculates ranking immediately after the profile is updated.
  • Runs profile matching entirely on-device without additional AI requests or model costs.

Interface Improvements

  • Refined profile-match badges in Dark mode with better colors, borders, and visual hierarchy for high, medium, and low matches.
  • Refined bidding-stage progress tracks and completed bars in Dark mode, replacing harsh white empty tracks with a subdued blue-gray treatment.

v0.9.5

Choose a tag to compare

@github-actions github-actions released this 26 Aug 03:10

中文说明

新增

  • 新增招投标「模板管理」二级功能,支持模板创建、实时预览、编辑、删除、导入和导出。
  • 支持配置纸张与页边距、封面、页眉页脚、一级至九级标题、正文、列表、表格和图片样式。
  • 标题支持 Word 自动多级编号;关闭自动编号后仍可按模板生成可手动调整的静态编号。
  • 新增表格题注并固定显示在表格上方,图片与表格题注均可独立配置。
  • 技术方案、已有方案扩写和可行性研究报告的 Word 导出支持选择基础格式、优化排版或自定义模板。
  • 导出模板支持按名称检索,并显示匹配数量,方便模板较多时快速选择。
  • 字体选择支持检索系统字体,并将常用中文字体置顶。

优化与修复

  • 完善经典、柔光和暗夜三种主题下的模板编辑、实时预览、字体选择、下拉框和全屏预览显示效果。
  • 优化模板列表、编辑按钮、全屏退出按钮及各类表单控件的圆角和视觉层级。
  • 修复暗夜模式下标题预览、自定义编号说明、无序列表、首行缩进及部分下拉框对比度异常。
  • 修复招投标相关模块 Word 导出弹窗显示异常及模板配置未正确应用的问题。
  • 模板设置仅作用于招投标相关模块,不影响其他业务模块的 Word 导出。

English Notes

Added

  • Added a dedicated Template Management section for bidding workflows, with create, live preview, edit, delete, import, and export support.
  • Added configuration for paper and margins, covers, headers and footers, heading levels 1–9, body text, lists, tables, and images.
  • Added native Word multilevel heading numbering. When automatic numbering is disabled, templates can still emit static numbers that remain manually editable in Word.
  • Added table captions above tables, with independent caption styling for both tables and images.
  • Added Word export choices for technical plans, existing-plan expansion, and feasibility reports: basic formatting, optimized formatting, or a saved custom template.
  • Added template-name search and match counts to the Word export dialog for faster selection in large template libraries.
  • Added searchable system-font selection with commonly used Chinese fonts pinned to the top.

Improvements and Fixes

  • Improved template editing, live preview, font selection, dropdowns, and fullscreen preview across Classic, Soft, and Dark themes.
  • Refined template cards, edit controls, fullscreen exit controls, form controls, corner radii, and visual hierarchy.
  • Fixed low-contrast heading previews, custom-numbering guidance, unordered-list symbols, first-line indentation fields, and dropdown states in Dark mode.
  • Fixed Word export dialog rendering and custom-template application across bidding workflows.
  • Scoped template settings to bidding workflows so Word exports in other modules remain unchanged.

v0.9.4

Choose a tag to compare

@github-actions github-actions released this 21 Aug 08:02

中文

  • 修复已知问题

English

  • Fixed known issues.

v0.9.2

Choose a tag to compare

@github-actions github-actions released this 21 Aug 07:04

V0.9.2

中文说明

新增功能

  • 在“招投标”下新增“可研报告”,提供项目创建与管理、项目资料录入、附件导入、资料分析、报告目录生成、关键参数整理、正文生成与校审、整篇预览和 Word 导出的一体化流程。
  • 可研报告目录支持模板、目标字数和知识库参考配置,并支持多级目录展开/折叠、编辑、增删和 AI 调整。
  • 正文支持按小节后台生成、暂停恢复、缺失补写、单节重生成、人工化校审及章节状态跟踪;页面切换不会中断任务。
  • 新增正文配图配置,可组合使用 AI 生图、技术图谱和 Mermaid 图,并在预览与 Word 导出中处理相关图片。
  • 新增可研报告项目重置、返回首页、暗黑模式适配和独立本地工作区存储。

修复与优化

  • 统一可研报告六个步骤的界面布局、进度反馈、目录树、编辑预览和底部操作栏样式,使其与技术方案流程保持一致。
  • 修复预览、全屏、目录设置、目录折叠、AI 调整、正文配图配置保存和 Word 导出等交互问题。
  • 改进生成与校审进度展示,避免任务完成或无待补写章节时进度停留在 1%,并在目录中标识正在校审的章节。
  • 修复暗黑模式下预览表头、更新下载进度区域及相关弹窗样式。
  • 修复 Word 导出中残留 Markdown 加粗标记的问题,并约束正文 Markdown 标题不再生成“一、”“(一)”或“(一)”等重复序号。
  • Word 导出改为底部功能按钮触发,增加导出进度弹窗、完成路径展示和打开文件目录能力。
  • 更新 Agnes AI 中国大陆站与国际站注册地址及跳转链接。

验证

  • 通过 Electron Main 服务语法检查。
  • 通过可研报告 Store 与后台任务验证脚本。
  • 通过 TypeScript 与 Vite 生产构建。

English Notes

New features

  • Added Feasibility Reports under Bidding, covering project management, project data entry, document import, source analysis, outline generation, key-parameter preparation, content generation and review, full-document preview, and Word export.
  • Added outline settings for templates, target word counts, and knowledge-base references, together with expandable multi-level trees, editing, insertion, deletion, and AI-assisted adjustment.
  • Added background section generation with pause/resume, missing-section completion, per-section regeneration, humanized review, and live chapter status tracking without interrupting tasks during page navigation.
  • Added configurable illustrations using AI-generated images, technical diagrams, and Mermaid diagrams, including preview and Word export handling.
  • Added project reset, return-to-home navigation, dark-mode support, and isolated local workspaces for feasibility-report projects.

Fixes and improvements

  • Unified all six feasibility-report steps with the technical-plan layout, including progress feedback, outline trees, editing/preview panels, and the floating action bar.
  • Fixed inactive preview, fullscreen, outline settings, tree folding, AI outline adjustment, illustration configuration saving, and Word export interactions.
  • Improved generation and review progress so completed or no-op completion tasks no longer remain at 1%, and reviewing chapters are visibly marked in the outline.
  • Fixed dark-theme colors for preview table headers, update-download progress, and related dialogs.
  • Fixed leaked Markdown bold markers in Word exports and constrained generated Markdown headings from adding duplicate Chinese ordinal prefixes such as “一、”, “(一)”, or “(一)”.
  • Moved Word export to the bottom action bar and added an export-progress dialog, completed file location display, and open-folder action.
  • Updated registration labels and links for the Agnes AI mainland China and international services.

Validation

  • Passed Electron Main service syntax checks.
  • Passed feasibility-report store and background-task verification scripts.
  • Passed the TypeScript and Vite production build.

Full Changelog: v0.9.1...v0.9.2

v0.9.1

Choose a tag to compare

@github-actions github-actions released this 20 Aug 09:04

V0.9.1

中文说明

修复与优化

  • 优化 Word 导出流程:保存位置确认前不再提前显示导出进度,减少暗黑模式下系统保存窗口与应用进度界面的视觉冲突。
  • 增强 Mermaid 本地渲染与 Word 导出兼容性,自动整理常见的不规范流程图、时序图语法,并修复宽图、高图被裁切或显示不完整的问题。
  • 将文本形式的流程图转换为图片后写入 Word,避免跨页后出现字符错位、线框断裂。
  • 改进专利交底书生成约束与导出校验,降低 AI 生成 Mermaid 语法导致本地渲染失败的概率。
  • 优化专利挖掘页面布局,减少结果区域不必要的留白。
  • 补齐暗黑模式样式,包括软件著作源码准备提示、公文模板标签及课题申报导出进度状态。
  • 更新应用 Logo 与 Windows、macOS 图标资源,并为图标保留统一圆角效果。

验证

  • 通过 TypeScript 与 Vite 生产构建。
  • 通过 Mermaid 本地渲染及 Word 嵌图回归验证。
  • 通过专利生成流程验证。

English Notes

Fixes and improvements

  • Improved the Word export flow so progress is shown only after the save location is confirmed, reducing visual conflicts between the native save dialog and the dark-themed application UI.
  • Hardened local Mermaid rendering and Word export compatibility by normalizing common malformed flowchart and sequence-diagram syntax and preventing wide or tall diagrams from being clipped.
  • Converted text-based diagrams into images before inserting them into Word to prevent broken borders and misaligned characters across pages.
  • Refined patent disclosure generation constraints and export validation to reduce local Mermaid rendering failures caused by AI-generated syntax.
  • Improved the patent-mining layout to remove unnecessary empty space in the results area.
  • Completed dark-theme styling for software-copyright source-preparation notices, official-document template tags, and grant-application export progress states.
  • Updated the application logo and Windows/macOS icon assets with consistent rounded corners.

Validation

  • Passed the TypeScript and Vite production build.
  • Passed local Mermaid rendering and Word image-embedding regression checks.
  • Passed the patent-generation workflow verification.

Full Changelog: v0.9.0...v0.9.1

v0.9.0

Choose a tag to compare

@github-actions github-actions released this 15 Aug 15:16

V0.9.0

中文说明

主要更新

  • 新增完整的暗黑主题,统一适配主页面、工作台、编辑器、弹窗、提示消息、加载状态和空状态,并减少模块切换时的闪白现象。
  • 优化售前项目与售前工作台,完善项目资料、客户分析、调研准备、方案架构、图表草稿和汇报页纲流程;页纲入口现改为只读详情预览,不再触发 PPT 导出。
  • 重构项目管理模块,完善项目类型、项目档案、阶段工作区、历史记录、导出模型和任务恢复能力。
  • 优化技术方案与已有方案扩写流程,统一目录、全局事实、正文生成、扩写、配置弹窗和原方案参考区域的交互与主题表现。
  • 完善公文写作、课题申报、论文导师、软著生成、专利生成、知识库、标书查重和投标机会等模块的界面与流程一致性。
  • 用量统计折线图新增数据点悬停提示,可查看对应日期和 Token 消耗量。
  • 拆分并统一跨进程类型契约与公共 UI,增强 IPC 校验、后台任务恢复和模块可维护性。

构建与体积

  • 精简 Mermaid 运行资源并增加打包校验,减少 Windows 与 macOS 安装包中的冗余文件。
  • 增加 IPC、资源打包、任务恢复、项目管理、论文导师和 Mermaid 导出的验证脚本。
  • 发布工作流继续自动生成 Windows 与 macOS 安装包;当前安装包未进行代码签名。

English Notes

Highlights

  • Added a complete dark theme across pages, workspaces, editors, dialogs, notifications, loading states, and empty states, while reducing white flashes during module transitions.
  • Improved presales project management and the presales workspace across project materials, customer analysis, research preparation, solution architecture, diagram drafts, and presentation outlines. The outline action is now a read-only detail preview and no longer triggers PPT export.
  • Refactored project management with improved project types, project profiles, stage workspaces, history, export models, and background task recovery.
  • Improved the technical plan and existing-plan expansion workflows, including outline generation, global facts, content generation, expansion, configuration dialogs, and original-plan references.
  • Polished UI and workflow consistency across official document writing, grant applications, thesis tutoring, software copyright, patent generation, knowledge base, duplicate checking, and bid opportunities.
  • Added per-point hover details to the Token usage trend chart, showing the date and Token consumption.
  • Consolidated shared process contracts and UI components, with stronger IPC validation, background task recovery, and module maintainability.

Build and Package Size

  • Reduced redundant Mermaid runtime assets and added packaging verification for smaller Windows and macOS distributions.
  • Added verification scripts for IPC contracts, packaged resources, task recovery, project management, thesis tutoring, and Mermaid export.
  • The release workflow continues to build Windows and macOS packages automatically. Packages are currently unsigned.

v0.8.7

Choose a tag to compare

@github-actions github-actions released this 13 Aug 02:17

禹都 AI 解决方案助手 v0.8.7

中文更新说明

本版本重点升级“投标机会”,形成从公开公告扫描、机会汇总、商机研判到售前与投标生产衔接的本地工作流,并持续优化桌面窗体的视觉与响应式体验。

投标机会总览与工作台

  • 新增投标机会总览,集中展示机会规模、预算池、截止风险、待办任务、投标推进、决策情况和最近扫描结果。
  • 新增机会详情工作台,支持收件箱筛选、状态管理、批量处理、负责人分配、投标决策流程和跟进记录。
  • 新增项目公告时间线,可识别采购意向、招标、更正、中标成交及废标终止等阶段,并支持人工合并或拆分项目链路。
  • 支持将正式招标文件流转到技术方案、废标项检查和售前工作台,减少重复导入与解析。

公告扫描与商机监测

  • 接入中国政府采购网中央及地方公开招标、采购意向、更正、中标成交和废标终止等公开公告来源。
  • 扫描仅在用户主动触发时执行,不绕过登录、验证码或网站访问限制,也不会使用演示数据替代失败结果。
  • 新增监控方案、关键词与地区条件、预算范围、采购人条件和公告类型匹配。
  • 新增雷达扫描效果、扫描进度、来源健康状态和具体异常诊断,支持直接定位并重新扫描异常来源。

研判、数据安全与可靠性

  • 新增企业能力画像、资格门槛、商机价值、投标准备、规则初筛和 AI 深度分析,并区分本地规则结果与 AI 结论。
  • AI 分析保留原文证据及人工核验状态;公告或企业画像变化后会提示重新分析。
  • 新增本地 SQLite 工作区、机会公告正文缓存、可验证备份、异常任务恢复和投标机会专项验证脚本。
  • 优化关联项目统计和首屏数据读取,减少进入投标机会页面时的等待时间。

界面与桌面适配

  • 重构投标机会总览和详情界面,优化编辑按钮、创建售前项目按钮、状态信息块和公告原文阅读字号。
  • 新增加载进度面板和居中雷达动效,替代大面积空白骨架屏。
  • 改进非最大化窗体下的响应式布局,三栏工作台、顶部操作栏和详情标签会根据可用内容宽度自动收缩或换行。
  • 暂时隐藏商务标入口,当前版本聚焦技术方案、机会发现和投标协同流程。

发布说明

  • macOS 与 Windows 安装包仍未进行代码签名,首次打开时可能出现操作系统安全提示。

English Release Notes

This release delivers a major upgrade to Bid Opportunities, adding a local workflow for public-notice scanning, opportunity overview, bid qualification, and handoff to presales and bid-production modules, together with continued desktop UI and responsive-layout improvements.

Opportunity Overview and Workspace

  • Added an opportunity overview covering pipeline size, budget, deadline risks, tasks, bid stages, decisions, and recent scan results.
  • Added a detailed opportunity workspace with inbox filters, status management, bulk actions, owner assignment, bid-decision workflows, and follow-up records.
  • Added project notice timelines for procurement intentions, tenders, corrections, awards, deals, terminations, and related stages, including manual merge and split operations.
  • Added handoff of formal tender documents to Technical Plans, Rejection Checks, and the Presales Workspace without repeated import and parsing.

Public Notice Scanning and Monitoring

  • Added public sources from the China Government Procurement Network for central and local tenders, procurement intentions, corrections, awards, deals, and terminated notices.
  • Scans run only when explicitly initiated by the user. The client does not bypass authentication, CAPTCHA, or access restrictions, and does not substitute demo data for failed requests.
  • Added monitoring rules for keywords, regions, budget ranges, buyers, and notice types.
  • Added a radar scan experience, scan progress, source health status, actionable diagnostics, and direct rescan actions for failed sources.

Qualification, Data Safety, and Reliability

  • Added enterprise capability profiles, qualification gates, opportunity value, bid readiness, rule-based screening, and AI-assisted deep analysis, while keeping local rule results separate from AI conclusions.
  • AI analysis retains source evidence and verification status, and becomes stale when the notice or enterprise profile changes.
  • Added a local SQLite workspace, notice-content cache, verifiable backups, interrupted-task recovery, and dedicated Bid Opportunity verification scripts.
  • Optimized project-relation statistics and initial data loading to reduce the delay when opening Bid Opportunities.

Interface and Desktop Responsiveness

  • Redesigned the opportunity overview and detail workspace, including edit actions, presales creation, lifecycle information blocks, and notice-content typography.
  • Replaced the large blank skeleton with a centered loading panel, progress indicator, and radar animation.
  • Improved non-maximized-window behavior: the three-column workspace, command bar, and detail tabs now adapt to the available content width.
  • Temporarily hid the Commercial Bid entry so this release can focus on technical plans, opportunity discovery, and bid collaboration.

Distribution Note

  • macOS and Windows packages remain unsigned, so operating-system security warnings may appear on first launch.

Full Changelog: v0.8.6...v0.8.7

v0.8.6

Choose a tag to compare

@github-actions github-actions released this 11 Aug 12:59

禹都 AI 解决方案助手 v0.8.6

中文更新说明

本版本重点修复 Mermaid 图在 Word 导出中的显示问题,并优化全局任务提示的位置与本地导出说明。

Mermaid 本地导出

  • Mermaid 图的预览与 Word 图片转换均在客户端本地完成,不再调用 mermaid.ink 等第三方图片转换服务。
  • 使用 Electron Chromium 渲染并截图 Mermaid 图,修复导出后节点变黑、框内文字消失、时序图文字与箭头错位等问题。
  • 针对流程图与时序图的不同 SVG 坐标系分别处理尺寸,避免长图被裁切或时序图导出为空白。
  • 修复连续导出多张 Mermaid 图时偶发空白的问题,并清理本地渲染窗口的加载监听器。
  • 增加 Mermaid 流程图、时序图及 SVG 样式转换回归验证,覆盖文字、节点颜色、箭头和连续导出场景。

界面与提示

  • 将全局任务运行提示移至内容区左上方,避免与右上角界面风格切换控件重叠。
  • 更新正文生成配置和 Word 导出进度文案,明确 Mermaid 转图在本地完成、无需联网。

发布说明

  • macOS 与 Windows 安装包仍未进行代码签名,首次打开时可能出现操作系统安全提示。

English Release Notes

This release focuses on fixing Mermaid rendering in Word exports, improving the global task indicator placement, and clarifying that Mermaid image conversion runs locally.

Local Mermaid Export

  • Mermaid preview and Word image conversion now run entirely on the local client without calling third-party conversion services such as mermaid.ink.
  • Mermaid diagrams are rendered and captured with Electron Chromium, fixing black nodes, missing node labels, and displaced text or arrows in sequence diagrams.
  • Flowcharts and sequence diagrams now use coordinate-aware sizing to prevent cropped long diagrams and blank sequence-diagram exports.
  • Fixed intermittent blank images when exporting multiple Mermaid diagrams consecutively and cleaned up render-window load listeners.
  • Added regression coverage for Mermaid flowcharts, sequence diagrams, SVG class styles, text visibility, node colors, arrows, and repeated exports.

Interface and Messaging

  • Moved the global running-task indicator to the upper-left of the content area so it no longer overlaps the theme-style switcher.
  • Updated content-generation and Word-export messages to state that Mermaid conversion is local and does not require network access.

Distribution Note

  • macOS and Windows packages remain unsigned, so operating-system security warnings may appear on first launch.

Full Changelog: v0.8.5...v0.8.6

v0.8.5

Choose a tag to compare

@github-actions github-actions released this 11 Aug 06:11

禹都 AI 解决方案助手 v0.8.5

中文更新说明

本版本重点升级软件著作权模块,形成从软著项目、源码准备、材料生成、人工核对、提交前总检到正式资料导出的完整工作流。

软著项目与源码准备

  • 新增独立的软著项目管理,支持创建、检索、进入和删除项目;删除操作提供二次确认。
  • 每个软著项目独立保存源码范围、登记字段、草稿、复核状态和导出记录,项目之间互不覆盖。
  • 将“源码准备”与当前软著项目关联,优化首次使用引导、文件筛选、重新扫描和材料生成入口。
  • 增强源码技术栈识别,结合项目配置、依赖清单、文件结构和语言分布识别常见前后端及桌面应用技术。

材料生成与编辑

  • 重构软著材料页面,按“源码与登记信息—草稿生成—材料核对—申报复核—提交前总检—正式导出”的顺序组织功能。
  • 支持使用文本模型根据源码证据生成软件主要功能和软件技术特点,并保留人工编辑与保存能力。
  • 登记日期改用日期选择组件,减少手工输入格式错误。
  • 新增草稿结构化审阅、章节编辑、预览、版本记录和恢复能力。
  • 改进操作手册标题一致性检查和问题定位,支持直接跳转到对应章节处理。
  • 优化代码鉴别材料的 DOCX 排版、分页、长行处理、字体和页眉页码显示。

操作手册图片与 AI 示意图

  • 支持纯文字手册、手动截图和 AI 功能示意图三种模式。
  • AI 可根据导入源码的功能证据生成多组相关提示词,并支持一键更换提示词。
  • 新增图片悬停预览、放大查看、说明与插入位置编辑、排序、移除和自定义提示词重新生成。
  • 新增操作手册图片人工核对,图片内容、说明或位置发生变化后会自动要求重新确认。

核对、总检与正式导出

  • 新增代码鉴别材料人工核对,覆盖页数范围、源码范围和代码可读性。
  • 新增申报人工复核与证据链清单,覆盖软件权属、申请主体、关键日期、源码证据和当地受理要求。
  • 新增官网字段映射、跨材料一致性检查、材料就绪清单和清晰的“去处理”定位入口。
  • 调整正式流程:提交前总检全部通过后才可导出正式资料;后台同样强制执行前置校验,不能绕过。
  • 将导出后的文件摘要、命名和 ZIP 完整性检查独立为“交付包校验”,不再与导出前总检形成循环。
  • 支持导出批次历史、确认快照绑定和文件完整性复查。

稳定性与发布

  • 增加软著工作区迁移、项目级存储、草稿历史和异常恢复保护。
  • 加固本地图片尺寸解析及文档图片处理,升级相关依赖并增加安全验证脚本。
  • 增加软著、源码准备、文档安全、发布候选版和未签名产物验证脚本。
  • macOS 与 Windows 版本暂不进行代码签名,首次打开时可能出现操作系统安全提示。

English Release Notes

This release delivers a major upgrade to the Software Copyright module, providing an end-to-end workflow from project and source preparation through drafting, human review, pre-export checks, and final package export.

Software Copyright Projects and Source Preparation

  • Added dedicated Software Copyright project management with create, search, open, and delete operations, including a confirmation dialog for deletion.
  • Each project now keeps its own source scope, registration fields, drafts, review state, and export history without overwriting other projects.
  • Connected Source Preparation to the active Software Copyright project and improved first-use guidance, file selection, rescanning, and handoff to material generation.
  • Improved technology-stack detection using project configuration files, dependency manifests, directory structure, and language distribution.

Material Generation and Editing

  • Reorganized the workspace into a clear sequence: source and registration data, draft generation, material review, filing review, pre-export checks, and final export.
  • Added text-model generation for the software's main functions and technical features based on source evidence, with editable and saveable results.
  • Replaced manually typed registration dates with date-picker controls.
  • Added structured draft review, chapter editing, preview, version history, and restore support.
  • Improved operation-manual title consistency checks and direct navigation to the relevant section.
  • Improved code-identification DOCX layout, pagination, long-line handling, fonts, headers, and page numbering.

Manual Images and AI Illustrations

  • Added three manual modes: text only, imported screenshots, and AI-generated functional illustrations.
  • The text model can now generate multiple source-related image prompts, with one-click prompt replacement.
  • Added hover preview, full-size viewing, caption and placement editing, reordering, removal, and custom-prompt regeneration.
  • Added human confirmation for operation-manual images; changes to content, captions, or placement automatically invalidate the previous confirmation.

Review, Pre-export Checks, and Final Export

  • Added human review for code-identification materials, covering page range, source scope, and readability.
  • Added a filing review and evidence-chain checklist covering ownership, applicant identity, key dates, source evidence, and local filing requirements.
  • Added official-form field mapping, cross-material consistency checks, a readiness checklist, and direct “Go fix” navigation.
  • Final export is now the last step and is enabled only after all pre-export checks pass. The Electron backend enforces the same gate so it cannot be bypassed.
  • Moved post-export hash, naming, and ZIP integrity checks into a separate Delivery Package Verification section.
  • Added export-batch history, confirmed-snapshot binding, and file-integrity revalidation.

Reliability and Distribution

  • Added workspace migration, project-scoped persistence, draft history, and recovery safeguards for Software Copyright workflows.
  • Hardened local image-dimension parsing and document image handling, upgraded related dependencies, and added security verification scripts.
  • Added verification scripts for Software Copyright, source preparation, document security, release candidates, artifacts, and unsigned packages.
  • macOS and Windows packages remain unsigned, so operating-system security warnings may appear on first launch.

Full Changelog: v0.8.0...v0.8.5

v0.8.0

Choose a tag to compare

@github-actions github-actions released this 07 Aug 06:10

禹都 AI 解决方案助手 v0.8.0

中文说明

本版本重点增强模型服务商兼容性、长任务稳定性、废标检查能力与本地使用体验。所有新增统计仍仅保存在本机,不接入远程埋点或远程 Analytics。

模型与服务商

  • 完善 Agnes AI 中国大陆站与国际站支持,内置 Agnes 2.0/2.5 Flash、2.5 Pro/Pro Alpha 及 Agnes Image 2.0/2.1 Flash。
  • 增强 DeepSeek 与 LongCat 模型配置,支持 Thinking 模式、推理强度和已知能力信息。
  • 新增模型能力探测、官方能力回退和本地缓存,模型列表仍由用户主动获取。
  • 优化生图模型质量、比例及兼容参数。

稳定性与安全

  • 增加统一的安全网络请求能力,包括超时、取消、响应大小限制和远程 URL 校验。
  • 加固 AI 生图、MinerU 上传下载、文档远程图片和 Word 导出图片链路。
  • 配置文件采用临时文件替换写入,降低异常退出造成配置损坏的风险。
  • 改进后台任务中断恢复、任务状态回放和全局任务监视器。
  • Mermaid 改为本地渲染与导出,提升稳定性并减少外部依赖。

投标与废标检查

  • 废标检查支持一次导入多份投标、资格、商务和技术文件。
  • 多文件合并分析时保留文件名,风险证据可以追溯到具体来源文件。
  • 新增“招标要求—投标响应—证据位置—状态—风险”符合性矩阵。
  • 优化响应文件要求提取、目录生成和后续正文生成的数据传递。

本地用量与界面

  • 新增本地 Token 用量统计,包括请求次数、输入、输出、Thinking 和总 Token。
  • 支持近 1 小时、6 小时、1 天、7 天和 14 天趋势曲线。
  • 新增模型用量排行、紧凑指标卡片和本地统计清理功能。
  • 调整设置页布局、Thinking Token 预算显示及提示样式。
  • 修复 Toast 生命周期与会话状态问题。

其他说明

  • 本版本不会新增远程统计、匿名标识或远程公告服务。
  • macOS 与 Windows 构建仍未接入代码签名,首次运行时可能出现系统安全提示。

English Release Notes

This release focuses on model-provider compatibility, long-running task reliability, bid compliance review, and local usability. All newly added usage statistics remain on the local device; no remote telemetry or analytics service is introduced.

Models and Providers

  • Improved Agnes AI support for Mainland China and international endpoints, with built-in presets for Agnes 2.0/2.5 Flash, 2.5 Pro/Pro Alpha, and Agnes Image 2.0/2.1 Flash.
  • Enhanced DeepSeek and LongCat configuration with Thinking mode, reasoning effort, and known capability metadata.
  • Added model capability probing, official capability fallback, and local caching. Remote model discovery remains user initiated.
  • Improved image-model quality, aspect-ratio, and compatibility parameters.

Reliability and Security

  • Added a shared secure HTTP layer with timeouts, cancellation, response-size limits, and remote URL validation.
  • Hardened AI image downloads, MinerU upload/download flows, remote document images, and Word export images.
  • Switched configuration persistence to temporary-file replacement to reduce corruption after unexpected shutdowns.
  • Improved interrupted-task recovery, task-state replay, and the global background-task monitor.
  • Moved Mermaid rendering and export to the local runtime for better reliability and fewer external dependencies.

Bidding and Compliance Review

  • Added multi-file import for bid, qualification, commercial, and technical documents.
  • Preserved source filenames during merged analysis so evidence can be traced to individual files.
  • Added a compliance matrix covering tender requirements, bid responses, evidence locations, status, and risk.
  • Improved response-file requirement extraction and propagation into outline and content generation.

Local Usage and UI

  • Added local Token usage statistics for requests, input, output, Thinking, and total Tokens.
  • Added trend ranges for the last 1 hour, 6 hours, 1 day, 7 days, and 14 days.
  • Added per-model rankings, compact metric cards, and local-statistics cleanup.
  • Improved settings layouts, Thinking Token budget presentation, and status messaging.
  • Fixed Toast lifecycle and session-state issues.

Notes

  • This release does not add remote telemetry, anonymous identifiers, or remote announcement services.
  • macOS and Windows builds are currently unsigned, so operating-system security warnings may appear on first launch.