Prepare desktop 1.15 offline source and release acceptance

This commit is contained in:
Agent Workbench Dev
2026-10-31 11:50:51 +08:00
parent b77783ba67
commit c5c8a84a7b
51 changed files with 6211 additions and 1136 deletions
+6
View File
@@ -22,3 +22,9 @@ docs/desktop-projects-*.txt binary
docs/desktop-projects-*.json -text whitespace=blank-at-eol,blank-at-eof,space-before-tab,cr-at-eol
docs/desktop-controlled-*.txt binary
docs/desktop-controlled-*.json -text whitespace=blank-at-eol,blank-at-eof,space-before-tab,cr-at-eol
docs/desktop-host-*.tap binary
docs/desktop-host-*.txt binary
docs/desktop-host-*.json -text whitespace=blank-at-eol,blank-at-eof,space-before-tab,cr-at-eol
docs/desktop-credentials-*.tap binary
docs/desktop-credentials-*.txt binary
docs/desktop-credentials-*.json -text whitespace=blank-at-eol,blank-at-eof,space-before-tab,cr-at-eol
+3 -1
View File
@@ -39,7 +39,9 @@
- 本仓库已启用本地Git,v1.0.0-baseline为已验证基线;各切片使用具体开发分支,1.12交付位于feat/model-center-1.12。
- 1.13已交付两个受限worker、当前成果/冲突与父工程审查、停止恢复和累计预算,代码e1a2661及QA补强91cdd45位于feat/isolated-workers-1.13。新三工程13节点、父Node8/8/React9/9/Python10/10、React两视口各10断言、完整UI21类160组与独立404/404通过;最终不可变20261030193314594归档/便携及正式1.13.0升级通过。原PID23172/Provider、2连接/1池/6既有请求摘要及当时16项目/28会话/2551记录/4568事件每行/全导出保持,部署后10页通过。原403/404失败保留;后续1.14实际状态以下方记录为准,真实模型工程继续暂缓。
- 1.14已交付固定DeepSeek Harness受控工具/模型桥、所属prompt/审批/预算/原结果/停止及父工程审查/Git,Purge两版本来源锁定提示实验默认关闭;产品53e196a、QA路径57d075e/便携Git b311116位于feat/controlled-engines-1.14。新三工程13节点、父8/8/9/9/10/10、React两视口各8断言、23类178组UI及独立445/445通过;最终不可变20261030235440461便携11427/源码9748文件归档/实际随包及正式1.14.0升级通过。原PID23172/Provider、2连接/1池/6原请求摘要及16项目/28会话/2551记录/4568事件、全部7表每行/全导出保持,部署后14页通过,9229关闭。升级后QA误判原Node路径/SQLite行原型的失败与原快照保留,续验未重放升级;当前宿主execPath仍原1.12启动包,真实二进制SHA与1.14完全相同,runtimeRoot才表示当前处理器。见docs/QA-1.14.md及发行身份;下一PLAN-1.15桌面/本机加密连接尚未实现,真实推理/Purge收益继续暂缓。
- 1.14已交付固定DeepSeek Harness受控工具/模型桥、所属prompt/审批/预算/原结果/停止及父工程审查/Git,Purge两版本来源锁定提示实验默认关闭;产品53e196a、QA路径57d075e/便携Git b311116位于feat/controlled-engines-1.14。新三工程13节点、父8/8/9/9/10/10、React两视口各8断言、23类178组UI及独立445/445通过;最终不可变20261030235440461便携11427/源码9748文件归档/实际随包及正式1.14.0升级通过。原PID23172/Provider、2连接/1池/6原请求摘要及16项目/28会话/2551记录/4568事件、全部7表每行/全导出保持,部署后14页通过,9229关闭。升级后QA误判原Node路径/SQLite行原型的失败与原快照保留,续验未重放升级;当前宿主execPath仍原1.12启动包,真实二进制SHA与1.14完全相同,runtimeRoot才表示当前处理器。见docs/QA-1.14.md及发行身份;1.14为正式历史基线;当前1.15进度见下方,真实推理/Purge收益继续暂缓。
- 1.15的A–E已提交249f191/72a496a/5354d95/f7a5969/b77783b:实际Windows宿主身份/单实例、CurrentUser DPAPI、Electron六入口、可回退安装/分页巡检及桌面三工程16节点/原生UI9组通过。F冻结1.15代码/完整回归/独立源码/最终安装/便携与保留状态正式部署正在执行;不能拿开发版1.14结果代替最终发行。离线源码使用随包Electron、npm ci --offline --ignore-scripts及Node24,核对全部dist-client/server/desktop。窗口关闭不结束执行宿主;旧内存Key不追溯保存。下一PLAN-1.16只计划,真实手机/多模态/其它引擎/推理收益分别待验。本轮不新建代理、goal、automation或其他chat。
- 用户授权后续按可审查小步本地提交。每个独立实现/修复/规划改动通过相关检查后,检查diff并提交,报告commit与结果;只读且无变化时不做空提交。
- 常规后续提交只暂存本次修改范围;保留其他人/用户的已有改动。
- 实验使用具体任务分支/worktree,合并前验证;不要用破坏性reset/clean丢弃工作。
+8
View File
@@ -1,5 +1,13 @@
# Agent Workbench 更新记录
## 1.15.0 · 2026-10-31(发行验收中)
- 增加实际Windows用户/进程创建/安装输入身份核对、签名挑战和并发单实例;桌面窗口与Node执行宿主分离,关闭重开保持原Task、Run、审批和完成结果。
- 新连接可选择CurrentUser DPAPI本机加密保存,与连接登记共同提交;保存失败恢复原行/路由/内存引用,关闭加载、删除密文与清除内存分别控制。
- 固定Electron45.2.0和六个preload入口,支持原生项目目录选择、所属文件/项目定位、当前Task审查与明确退出宿主;界面沿用相同执行内核。
- 用户级安装使用独立versions目录、current.json和原生启动器;完整清单/活动守卫、根入口及指针一起回退,更新保留全部持久行/工件,卸载保留用户数据。运行巡检可分页核对更早的原工件,显示实际检查范围。
- A–E专项和实际安装/桌面三工程通过,原失败保留;F独立源码/全部历史UI/最终安装/便携及保留状态正式升级正在验收。正式服务目前仍1.14.0,完整状态见[QA-1.15](docs/QA-1.15.md),下一手机方案见[PLAN-1.16](docs/PLAN-1.16.md)。
## 1.14.0 · 2026-10-31
- 固定已有DeepSeek Harness及独立profile,禁用原生项目/终端/模型插件;原prompt、工具和模型请求进入工作台所属任务、审批、预算与原工件账本,成果经父工程审查、检查及Git交付。
+7 -5
View File
@@ -2,6 +2,8 @@
Agent Workbench 将 Claude Code 与 Cursor 的机制研究整合为独立的本地 AI 编程工作台,并内置模型池、项目记忆、Skills/MCP、REA 工件研究、受控DeepSeek Harness工程引擎、实际页面验收和有依赖的执行计划。工程索引、精确补丁、多文件整批审查、Node/Python实际检查、浏览器交互、本地Git提交及交付报告可以在同一个Task中完成。
1.15.0桌面阶段已实现并通过A–E专项验收:Windows宿主身份/单实例、可选择的本机加密连接、Electron窗口与本机项目入口、安装/更新/回退/卸载、分页历史工件巡检,以及桌面Node/React/Python工程交付。窗口关闭后后台宿主继续保留任务,重开恢复同一记录。最终独立发行与正式部署正在F验收,当前正式服务仍为1.14.0;见[1.15验收与实际范围](docs/QA-1.15.md)、[桌面使用说明](docs/DESKTOP.md)。
1.14.0已正式本地交付固定版本的受控引擎与工具/模型桥,原生prompt、审批、预算、原结果和停止进入所属Task/Run;独立工作区成果由父工程审查、检查及Git交付。Purge两版本来源锁定的提示实验默认关闭。三个真实功能工程13节点、23类178组UI、独立源码445/445及最终源码/便携验收通过;原PID、模型连接和全部历史保持,部署后14页复验通过。当前[本地工作台](http://127.0.0.1:4318/?release=1.14.0)、发行身份与具体结果见[QA-1.14](docs/QA-1.14.md)。真实模型推理收益按原选择独立待验。
1.13.0已本地交付[隔离协作](docs/WORKERS.md):父目标与两个受限Git工作区、累计预算/停止恢复、当前成果核对、具体冲突和父工程审查/检查/Git交付。独立源码404/404、完整界面21类160组与三个真实功能工程通过,父Node8/8、React9/9、Python10/10及构建/编译/Git,React内置Edge两视口各10项断言。正式[本地工作台](http://127.0.0.1:4318/?release=1.13.0)已保持原进程/模型连接/全部历史升级,验收与部署见[QA-1.13](docs/QA-1.13.md)。
@@ -22,7 +24,7 @@ Agent Workbench 将 Claude Code 与 Cursor 的机制研究整合为独立的本
## 后续产品与能力建设
1.12包含模型中心/池路由、[项目记忆](docs/PROJECT-MEMORY.md)、[任务恢复检查点](docs/TASK-CHECKPOINTS.md)、[Skills](docs/SKILLS.md)和[项目MCP](docs/MCP.md)。记忆只在来源复核、脱敏且请求预算允许时使用;恢复检查点绑定完整要求和原结果,核对后显式续接同一任务。1.13两个受限worker与1.14受控引擎/提示实验均已完成发行;下一阶段按[1.15桌面方案](docs/PLAN-1.15.md)开发窗口与宿主分离、本机加密连接、单实例及可回退安装,再推进真实手机配对、更多模型协议和多模态。
1.12包含模型中心/池路由、[项目记忆](docs/PROJECT-MEMORY.md)、[任务恢复检查点](docs/TASK-CHECKPOINTS.md)、[Skills](docs/SKILLS.md)和[项目MCP](docs/MCP.md)。记忆只在来源复核、脱敏且请求预算允许时使用;恢复检查点绑定完整要求和原结果,核对后显式续接同一任务。1.13两个受限worker与1.14受控引擎/提示实验均已完成发行;1.15桌面和本机加密A–E已验收,最终发行/部署以[QA-1.15](docs/QA-1.15.md)为准;下一[1.16手机伴随端](docs/PLAN-1.16.md)先实现一次性配对、设备权限与断线恢复,再推进更多模型协议和多模态。
当前Git项目可以先预览已提交版本,再创建独立目录和分支;来源已有暂存、未提交及未跟踪文件保留在原项目。子工作区有自己的项目、会话和实际执行目录,创建中断后显式核对。模型准备Git交付预览后停止请求,确认这份预览后由内核完成提交和报告。具体范围见[独立工作区](docs/WORKSPACES.md)及[1.6验收](docs/QA-1.6.md)。
@@ -39,7 +41,7 @@ Agent Workbench 将 Claude Code 与 Cursor 的机制研究整合为独立的本
## 直接运行
Windows x64 便携包自带 Node.js 24.22.1、npm、REA 6.8.0、Playwright 1.64.1及锁定生产依赖。页面验收使用本机已安装的Microsoft Edge。解压后双击 `Start-Workbench.cmd`,浏览器打开 `http://127.0.0.1:4318`。双击 `Stop-Workbench.cmd` 可停止服务及它正在运行的任务。
Windows x64 1.15桌面/便携包自带Node.js 24.22.1、npm、Electron 45.2.0、REA 6.8.0、Playwright 1.64.1及锁定生产依赖。页面验收使用本机Microsoft Edge。安装版双击安装目录的AgentWorkbench.exe,便携版双击Start-Desktop.cmd打开桌面;Start-Workbench.cmd仍可打开浏览器入口http://127.0.0.1:4318。最终1.15安装/发行状态见[QA-1.15](docs/QA-1.15.md)。
启动不需要安装 npm 依赖或管理员权限。窗口关闭后本地服务仍运行,任务与会话继续保存。再次打开相同地址即可查看记录。
@@ -65,7 +67,7 @@ REA首批为静态分析,研究工件不会自动运行。原结果与项目
模型网络故障、DNS、证书、超时、身份、限额和接口响应错误分别记录。可恢复请求最多尝试三次,尊重服务的 Retry-After,并限制总时长;401/403 和无效或过大的 JSON 不自动重试。停止任务会取消请求和退避。错误卡片的「重试模型请求」沿用已完成工具结果。
API Key 只在当前服务内存中使用,不写入配置、会话或导出。重启服务后需重新填写;也可在启动前设置专用环境变量 `AGENT_WORKBENCH_API_KEY`。工作台不读取其他应用的登录信息。
API Key默认只在执行宿主内存中使用。1.15可在「模型中心」对新填写或本次确认导入的连接选择本机加密保存,使用当前Windows用户的DPAPI密文,与连接登记同一SQLite事务提交;可关闭启动加载、删除本机保存或单独清除内存。临时连接重启后需重新填写。既有内存Key不会因升级自动写盘,公开配置、会话和导出不包含明文Key。格式导入只处理用户明确选定的来源,见[桌面与连接说明](docs/DESKTOP.md)。
接口契约已由本地 HTTP 测试服务及浏览器验证。用户随后配置的 `gpt-6.1-sol` 已通过真实读取、提案与应用验证,见[模型连接实测](docs/MODEL-CONNECTION.md)。
@@ -95,10 +97,10 @@ Windows 终端使用 PowerShell 5.1,通过 UTF-8 stdin 传入经确认的完
## 源码开发
使用 Node.js 24;依赖版本由 `package-lock.json` 固定。
使用 Node.js 24;依赖版本由 `package-lock.json` 固定。 1.15源码包已包含固定Electron和Node运行依赖,离线安装用`npm.cmd ci --offline --ignore-scripts`,构建后`npm.cmd run desktop`;源码仓库首次开发需先取得固定Electron 45.2.0发行运行文件。
```powershell
npm.cmd ci
npm.cmd ci --ignore-scripts
npm.cmd run build
npm.cmd test
```
+9 -7
View File
@@ -1,9 +1,11 @@
# Agent Workbench 能力目标与验收矩阵
日期:2026-10-30。1.0.0为基础里程碑,1.1.0实现M1,1.2.0增加内置Git交付,1.3.0增加工程索引和精确补丁,1.4.0增加可恢复多文件变更集,1.5.0/1.5.1增加有界请求与可用工件分页,1.6.0增加独立Git工作区和审查收尾,1.6.1增加实际Task工作依据和文件预算分页。下表保留1.0.0缺口与完整产品目标;每项完成范围以实际验证记录为准。
日期:2026-10-31。1.0.0为基础里程碑,1.1.0实现M1,1.2.0增加内置Git交付,1.3.0增加工程索引和精确补丁,1.4.0增加可恢复多文件变更集,1.5.0/1.5.1增加有界请求与可用工件分页,1.6.0增加独立Git工作区和审查收尾,1.6.1增加实际Task工作依据和文件预算分页。下表保留1.0.0缺口与完整产品目标;每项完成范围以实际验证记录为准。
总体方案见 [PRODUCT-PLAN.md](PRODUCT-PLAN.md),研究问题与来源见 [RESEARCH-ROADMAP.md](RESEARCH-ROADMAP.md)。
1.15当前实现与验收:A–E Windows宿主/单实例、DPAPI选择保存、真实Electron窗口、安装/回退/卸载和分页巡检、桌面三工程16节点及原生UI9组已通过并提交。F独立源码/最终发行/正式部署进行中,未计完整交付;原正式1.14服务和连接保持。具体证据见[QA-1.15](QA-1.15.md),下一真实手机前置见[PLAN-1.16](PLAN-1.16.md)。
2026-10-31当前正式交付基线为1.14.0:固定DeepSeek Harness受控工具/模型桥、单个所属worker、真实prompt/审批/检查/父审查/Git与停止核对,以及两版本来源锁定的默认关闭提示实验已交付。独立445/445、253份输入/6构建/40基准、三工程13节点、23类178组UI及最终便携/归档、保留状态升级通过。原PID23172/Provider、完整模型中心和全部7表每行/28会话导出保持,部署后14页通过。真实模型推理收益及其它引擎适配继续独立待验;见[1.14验收](QA-1.14.md)与[下一桌面阶段](PLAN-1.15.md)。
1.13历史交付:1.12.1逐里程碑/24项能力审查与五类确认问题修复保持,本阶段两个受限worker、父成果审查/恢复、独立404/404、完整160组UI、三工程13节点与最终源码/便携正式升级通过。原PID23172/Provider、完整模型中心及当时16项目/28会话/2551记录/4568事件每行和全导出保持,部署后10页通过,见[1.13验收](QA-1.13.md)和[回顾](RETROSPECTIVE-1.13.md)。各版本旧数字保留为历史证据,下一切片与未交付范围统一见[NEXT-STEPS](NEXT-STEPS.md)。
@@ -117,9 +119,9 @@
- C12/C13:成功准备Git预览后保存审查边界并停止模型;后续同消息工具不执行且协议完整。确认后内核完成原Task与报告,不继续请求模型,预览可修改/撤回。
- 192项回归和72组界面通过;合成模型子工作区真实3项失败→3项通过/build/Git报告通过。不可变包、原进程更新与真实配置模型结果见[QA-1.6](QA-1.6.md),完整三类工程及M2/M6能力继续验收。
| ID / 能力 | 当前已交付能力与缺口(至1.12.0) | 后续产品目标 | 参考问题 / 模块 / 阶段 | 必须看到的验收证据 |
| ID / 能力 | 当前能力与缺口(至1.15实现;发行状态见QA) | 后续产品目标 | 参考问题 / 模块 / 阶段 | 必须看到的验收证据 |
| --- | --- | --- | --- | --- |
| C01 需求与任务规划 | Task固定目标/追加要求、Run身份和不可变依赖步骤/真实证明已交付;跨worker任务调度待做 | 固定目标、约束和完成标准;拆任务、判断依赖、接收追加要求与重规划 | CX-1、WB-1 / tasks / M1–M2 | 追加要求后仍保留原目标;任务完成由验收条件决定 |
| C01 需求与任务规划 | Task固定目标/追加要求、Run身份、依赖步骤/真实证明及两个受限worker已交付;任意规模调度待做 | 固定目标、约束和完成标准;拆任务、判断依赖、接收追加要求与重规划 | CX-1、WB-1 / tasks / M1–M2 | 追加要求后仍保留原目标;任务完成由验收条件决定 |
| C02 执行状态与恢复 | SQLite事件/状态、原工件、Task检查点、unknown核对及显式同Task续接已交付;不重放已完成副作用 | 独立Task/Run/Step身份、持久事件、检查点与未知结果核对 | CC-2、CU-1、DS-1 / storage、tasks / M1 | 在审批、调用、结果与落盘等切点中断;恢复不重复已完成副作用 |
| C03 工程识别与索引 | 项目识别、JS/TS/React声明/导入索引与增量核对、Python候选已交付;完整语义/运行时依赖待做 | 识别语言、入口、规则、脚本、依赖和Git;增量索引;符号/引用与必要的语义检索 | CU-2、DS-3 / context / M2 | 对Node/React/Python样本定位实际修改入口;代码变动后索引更新 |
| C04 上下文与项目归属 | 有界原请求、工作依据/分页、实际工具反馈、来源记忆及Task/Run归属已交付 | 为Run组合需求、规则、代码、错误与验证结果;引用来源;严格绑定项目 | CC-2、CU-1 / context、contracts / M1–M2 | 切换两个项目/任务后,错误与工具结果仍归属正确;大输出用工件引用 |
@@ -136,12 +138,12 @@
| C15 预算与用量 | 完整请求预检、Run/Task累计预算与unknown usage已交付;1.13增加父总预算/两个子分配、累计步数/token/工具/时限、迟到和未知请求保守记账;价格与实际费用待做 | 时间、token、费用、步骤、重试与worker总预算;无进展检测;可配置价格与未知usage标记 | CC-2、DS-2 / models、tasks / M1–M4 | 子任务/重试纳入总预算;耗尽时保存检查点;缺失usage不伪造准确费用 |
| C16 多代理协作 | 1.13已交付两个受限worker、独立worktree、父子累计预算/停止恢复、当前成果与具体冲突、父工程审查合并;任意规模/OS沙箱待做 | 独立问题分工、上下文/预算隔离、独立worktree、结果与证据汇总、冲突合并 | CC-1、CU-3、DS-2、WB-1 / tasks、workspaces / M4 | 两个独立任务并行完成可合并;重叠编辑被发现;父任务停止可传播 |
| C17 REA能力研究 | 固定REA6.8.0的ASAR/PE三项静态操作、所属job/目标SHA256/原Evidence已交付;其余工具与运行trace待做 | 目标清单、版本/hash、inventory/analyze/trace、分析job、Evidence查看与研究结论登记 | 全部目标 / tools、artifacts / M3 | 从能力结论能追到确切工件与Evidence;静态/运行证据和未知范围分别显示 |
| C18 Purge增强实验 | DSH-purge来源/机制研究已登记,未集成工程执行或验证收益;兼容组合与同题对照待做 | 独立实验profile、兼容版本、策略分类、开关/回退和同题对照;有效机制按接口采纳 | PG-1 / engines、models、tasks / M3 | 同模型/任务/预算重复比较成功率、成本与错误修改;插件失效可恢复基线 |
| C19 外部代理引擎 | 固定DeepSeek0.2.1-rc.2的六方法连接/所属进程/原协议已交付;Codex适配及引擎prompt/工程写入待做 | EngineAdapter管理握手、会话/进程、事件、停止/恢复和真实能力限制 | CX-2、DS-1 / engines / M3 | 引擎事件有Task/Run映射;协议缺失能力明确;root idle不直接判工程完成 |
| C18 Purge增强实验 | 1.14来源锁定的两版本提示实验、开关/失效拒绝和同题夹具对照已交付;原完整插件与真实模型收益未验证 | 独立实验profile、兼容版本、策略分类、开关/回退和同题对照;有效机制按接口采纳 | PG-1 / engines、models、tasks / M3 | 同模型/任务/预算重复比较成功率、成本与错误修改;插件失效可恢复基线 |
| C19 外部代理引擎 | 1.14固定DeepSeek0.2.1-rc.2受控prompt/工具模型桥、所属工程/审批/预算/父审查/Git与停止已交付;Codex等适配待做 | EngineAdapter管理握手、会话/进程、事件、停止/恢复和真实能力限制 | CX-2、DS-1 / engines / M3 | 引擎事件有Task/Run映射;协议缺失能力明确;root idle不直接判工程完成 |
| C20 浏览器与界面验证 | 代理固定静态构建、真实Edge交互/DOM/PNG/原报告与Task绑定已交付;任意在线浏览与电脑操作待做 | 可选浏览器工具、页面状态/截图、请求/控制台结果、交互验收与开发预览 | CU-2、WB-1 / tools、artifacts / M3–M6 | React真实交互/失败场景可复现;结果含页面与任务身份,不只保存截图 |
| C21 桌面产品 | Windows浏览器便携/源码包及保留原进程升级已交付;Electron安装版、后台宿主/窗口拆分待做 | Electron Windows安装/升级、后台执行宿主、工程/终端/diff/成果区和任务通知 | 产品方案 / desktop、gateway / M5 | 关闭窗口后可再查看任务;安装/升级保留数据;宿主进程生命周期明确 |
| C21 桌面产品 | 1.15真实Electron窗口/后台宿主拆分、单实例、原生入口、安装/更新/回退/卸载A–E已验收;最终独立发行和正式部署待F | Electron Windows安装/升级、后台执行宿主、工程/终端/diff/成果区和任务通知 | 产品方案 / desktop、gateway / M5 | 关闭窗口后可再查看任务;安装/升级保留数据;宿主进程生命周期明确 |
| C22 手机伴随端 | 390px响应式界面已验证;电脑仅绑定127.0.0.1,真实手机PWA配对/撤销待做 | PWA任务/进度/审查入口;配对网关与设备撤销;共用Run与事件游标 | 产品方案 / mobile、gateway / M5 | 实际手机连电脑并审查同一任务;断线补发不重复执行;失效设备不能授权 |
| C23 数据、配置与发布迁移 | Key仅内存、SQLite/旧JSON迁移、不可变包/保留Provider与历史升级已交付;本机加密凭据、完整回退和设备配置待做 | 用户主动配置凭据、系统凭据存储、数据库迁移/回退、版本兼容和日志导出 | CX-2、DS-1 / storage、gateway、desktop / M1、M5–M6 | 升级副本与回退可验证;导出无Key;任务和来源不丢失 |
| C23 数据、配置与发布迁移 | 临时Key默认内存;1.15新增显式DPAPI保存/关闭/删除、同事务回退和安装根入口/指针回退已实际验收;正式发行待F,设备配置待1.16 | 用户主动配置凭据、系统凭据存储、数据库迁移/回退、版本兼容和日志导出 | CX-2、DS-1 / storage、gateway、desktop / M1、M5–M6 | 升级副本与回退可验证;导出无Key;任务和来源不丢失 |
| C24 工程能力评测 | 行为/UI与Node/React/Python实际样本、失败/人工介入证据已保留;封存成套基准、重复评测与整体稳定完成率待做 | 封存真实工程任务、独立验收、重复运行和错误分类;追踪人工介入/成本/恢复 | 全部目标 / evaluation、artifacts / M1–M6 | 发布能力可链接实际任务记录;报告含失败样本与环境,不能只展示成功演示 |
### 1.6.1 M2实际工作依据切片
+19
View File
@@ -0,0 +1,19 @@
# Windows桌面与本机连接
1.15复用现有工作台客户端和执行内核。A–E实际验收通过,最终发行/正式部署状态以[QA-1.15](QA-1.15.md)为准。
安装版默认为当前Windows用户安装在`%LOCALAPPDATA%\Programs\AgentWorkbench`;用户数据在`%LOCALAPPDATA%\AgentWorkbench`。安装目录含原生`AgentWorkbench.exe`入口、`current.json`和独立`versions`程序目录。新版本写入新目录,核对完整文件清单、原执行宿主和活动后才激活;失败回退当前指针和根入口。卸载先结束所属空闲宿主,用户数据保留。
双击安装目录的`AgentWorkbench.exe`打开桌面。便携包双击`Start-Desktop.cmd`;`Start-Workbench.cmd`仍可打开浏览器界面。源码包使用随包Node24、`npm ci --offline --ignore-scripts`和`npm run build`后,`npm run desktop`使用固定Electron运行目录。程序/data路径在首次自定义安装后按原登记复用,不能临时把另一份用户数据目录当原宿主。
普通关闭桌面窗口会退出该窗口,后台Node继续保留任务、模型连接和执行记录;重开核对实际用户、进程创建、安装输入、数据目录与签名挑战后复用所属宿主。同版本的其它目录也有独立身份,端口占用不自动当作本程序。需要结束执行时使用「退出执行宿主」,核对当前状态后等待所属工具、检查和原生引擎结束。已完成操作不会因重开窗口自动重放,未知结果仍需在原任务核对。
桌面提供原生项目目录选择、当前项目/文件定位和所属Task审查入口。文件入口核对已登记项目、实际路径和原SHA;不会把任意渲染页参数转为系统命令。六个固定preload方法运行在隔离窗口中,模型和客户端仍通过原Task/Run、文件审查与逐次命令审批执行。
模型Key默认只存于执行宿主内存。新填写或本次确认导入时,可选「本机加密保存」:当前Windows用户的DPAPI密文与连接登记共同提交。临时连接在宿主重启后显示待配置;保存连接可关闭启动自动加载、删除本机保存,或单独清除内存。关闭连接调度与删除密文是不同操作,界面分别显示状态。旧内存连接不因安装升级自动持久化,公开配置/会话/报告无明文Key。
密文绑定当前用户、数据目录、连接及配置修订;来源变化、损坏或解密失败会显示不能加载并保留原保存块。不同Windows账号登录尚未实际验收,错误SID/entropy/损坏和自有凭据实际重启/HTTP认证已分别验证。不要手动复制密文到另一数据目录作为连接迁移。
运行中心巡检为只读操作,不发送模型请求或运行项目命令。报告核对实际宿主/来源、SQLite、磁盘、所属worker/driver/MCP、未知结果和工件范围。历史工件按原页/游标分段核对,最近50份正常不等于全部历史正常;报告保留本次检查数和未覆盖数。发现损坏或unknown后仍需按原Task/Run核对,巡检不会解除状态或重放工具。
真实手机接入按[1.16方案](PLAN-1.16.md)实施一次性配对、设备权限、HTTPS与断线恢复。390px窄屏通过仅说明布局兼容,手机`127.0.0.1`指手机本身。真实推理/Purge收益、其它引擎、语音/视觉与通用电脑操控各有独立开发和验收门槛。
+4 -4
View File
@@ -1,10 +1,10 @@
# 后续执行计划
更新:2026-10-31。当前正式交付基线为1.14.0,[1.12.1里程碑审查](MILESTONE-AUDIT.md)及1.13隔离协作保持;[1.14受控引擎与提示实验](QA-1.14.md)的A–F、独立445/445、23类178组UI、三工程13节点、最终归档/便携和保留状态升级已完成。原PID23172/Provider、模型中心及16项目/28会话/2551记录/4568事件和全部7表每行/全部导出保持,部署后14页通过。下一阶段按[1.15方案](PLAN-1.15.md)开发Windows桌面与本机加密连接;真实模型工程及Purge推理收益仍按用户选择暂缓。历史研究、局部测试与连接核验不能代替真实工程交付。其他项目不进入本计划。
更新:2026-10-31。当前正式服务仍1.14.0,原PID23172/Provider、连接和全部历史保持。1.15 Windows桌面、持久连接与安装的A–E已完成实现/实际验收/本地提交,F独立发行与正式部署正在执行;见[QA-1.15](QA-1.15.md)。下一[1.16手机伴随端](PLAN-1.16.md)已记录具体配对、设备权限、真实手机、审批/停止、断线/撤销和独立发行门槛,须先完成1.15 F。真实模型工程/Purge推理收益继续按原选择暂缓,其他项目不进入本计划。
产品目标是让用户给出需求后,工作台能保留项目基线,完成定位、计划、修改、检查、必要修复、审查和Git交付,并在停止、故障、断线后明确保留已完成结果。桌面与手机共享同一执行宿主、任务和证据。
当前接续:1.14已正式交付,原失败、独立源码/随包/界面与升级证据均封存。下一阶段[Windows桌面与本机加密连接方案](PLAN-1.15.md)已记录六个切片及真实安装/重启/工程/回退标准,前置已通过;从A宿主身份、单实例与窗口生命周期开始,尚未实现桌面壳或持久凭据。
当前接续:完成1.15最终源码/历史UI/归档/便携/实际安装验收,以已验证候选安装保留状态升级正式4318,核对全部持久行/导出/工件和连接,再封存发行身份/回顾。只有这些完成后才进入1.16 A;计划不能代替真实手机结果。
## 执行顺序与完成门槛
@@ -14,7 +14,7 @@
| 1:隔离执行与多worker(完成) | 父任务拆分、两个所属worker、独立worktree、累计预算、停止恢复、成果归集、具体冲突及父工程审查 | 当前Task/Run、工作区、计划和未知结果恢复契约 | 三工程13节点、父Node8/8、React9/9、Python10/10与构建/编译/Git、React两视口各10断言、独立404项、160组UI、最终归档/便携与保留状态升级通过,见QA-1.13 |
| 2a:受控DeepSeek与提示对照(1.14完成) | 固定Harness/工具模型桥、所属工程/停止/父审查、两版本来源锁定默认关闭提示实验 | 外部SDK实际契约和1.13隔离工作区;运行器及实际输入身份固定 | 三工程13节点、独立445项、178组UI、实际随包/状态升级及14页通过;八工程36节点的夹具提示对照保留原输入/用量,真实推理收益未测,见QA-1.14 |
| 2b:其它引擎与真实收益(后续独立切片) | Codex实际适配;Claude Code/Cursor/WorkBuddy机制研究;Purge真实模型对照 | 实际版本、审批/取消/模型与工具契约;已验证的受限桥和所属工作区 | 指定工程有实际原协议/工具/检查/Git与停止结果;真实有无实验同模型同题同预算,记录失败和人工介入,不用连接或固定回答夹具推断成功率 |
| 3:Windows桌面 | 单实例、后台宿主、项目/终端/审查入口、安装包与保留数据更新 | 宿主与客户端生命周期拆分;本机凭据加密;稳定迁移契约 | 关闭/重开窗口不丢任务;重复启动复用宿主;安装/升级/卸载的数据行为可核验;崩溃后显式恢复;旧版本可回退 |
| 3:Windows桌面(A–E通过,F验收中) | 单实例、后台宿主、项目/终端/审查入口、安装包与保留数据更新 | 宿主与客户端生命周期拆分;本机凭据加密;稳定迁移契约 | 关闭/重开窗口不丢任务;重复启动复用宿主;安装/升级/卸载的数据行为可核验;崩溃后显式恢复;旧版本可回退 |
| 4:手机伴随端 | PWA、一次性配对、设备身份、撤销、同Run观察/审批、断线游标补发 | 设备认证、加密传输、审批身份/幂等、速率限制;阶段3宿主 | 在真实手机完成配对和撤销;断线后按seq恢复;旧Run/过期/重复审批拒绝;手机和桌面显示同一结果;手机不直接执行宿主文件与命令 |
| 5:模型与多模态 | 更多实际协议、流式取消、按能力选路、可恢复池策略、视觉/语音/附件、项目缓存 | 每种协议的工具调用/用量/取消实测;凭据和附件隔离 | 图像、语音、文本各有真实样本;流式断开保留unknown而不重复副作用;不兼容成员不被调度;缓存绑定项目/权限/源码版本并可清除;429/401/超时/冷却有独立证据 |
| 6:自我迭代与专业交付 | 隔离分支生成候选、固定评测集、多模型复核、质量/成本对照、升级候选和回滚 | 可复现工程基准、完整质量门槛、阶段1隔离和阶段3更新 | 工作台可提出并验证自身改进;候选不直接改运行中的发行;回归或独立验收失败不能发布;保存原差异、指标和回退点;采纳须经用户已有权限规则 |
@@ -37,7 +37,7 @@
按[PLAN-1.14](PLAN-1.14.md)已固定DeepSeek Harness的prompt/事件/审批/停止契约并通过受限桥完成所属工程、父审查和Git;实际SDK仍缺少session/cancel,停止单位为所属进程树,旧SDK会话不自动恢复,逐token流未实现。Codex等其它适配按具体版本独立推进,不能从连接成功推断能力。
DSH-purge两版本的来源锁定和同题本地HTTP夹具对照已完成,独立提示实验默认关闭;真实推理收益继续待验。下一[1.15桌面阶段](PLAN-1.15.md)先实现实际宿主身份/单实例,再本机加密连接、Electron窄范围窗口、可回退安装、实际桌面工程和独立发行。手机、跨协议多模态和自我迭代依照上表前置推进,每个切片继续回顾并本地Git。
DSH-purge两版本的来源锁定和同题本地HTTP夹具对照已完成,独立提示实验默认关闭;真实推理收益继续待验。当前[1.15桌面阶段](PLAN-1.15.md)A–E通过,F正执行独立发行/正式保留状态部署;下一[1.16](PLAN-1.16.md)按实际HTTPS和一次性配对开始。手机、跨协议多模态和自我迭代依照上表前置推进,每个切片继续回顾并本地Git。
## 跨阶段必须保持的能力
+1 -1
View File
@@ -1,6 +1,6 @@
# 1.15 Windows 桌面宿主与持久连接执行方案
2026-10-31。A–C已分别实现、验收和提交(249f191、72a496a、5354d95);D的16项补强回归、6组UI、11组实际安装演练通过并提交f7a5969;E的桌面三工程16节点和原生停止/恢复9组通过;F独立发行和正式部署尚未完成。1.14原445/445、178组UI、三工程及发行证据保持,具体前置结果见[1.14验收](QA-1.14.md)。每个切片保留原失败、回顾、复验并本地Git提交,失败的候选不进入正式服务。
2026-10-31。A–C已分别实现、验收和提交(249f191、72a496a、5354d95);D的16项补强回归、6组UI、11组实际安装演练通过并提交f7a5969;E的桌面三工程16节点和原生停止/恢复9组通过并提交b77783b;F元数据/源码离线运行与17相关检查已补齐,独立发行和正式部署进行中。1.14原445/445、178组UI、三工程及发行证据保持,具体前置结果见[1.14验收](QA-1.14.md)。每个切片保留原失败、回顾、复验并本地Git提交,失败的候选不进入正式服务。
## 本阶段交付目标
+42
View File
@@ -0,0 +1,42 @@
# 1.16 手机伴随端执行方案
2026-10-31记录。前置是[1.15](PLAN-1.15.md)完成F独立发行、正式桌面安装与保留状态部署;本文为下一阶段计划,不能据此认为手机功能已经实现。
## 用户成果
用户在Windows工作台开启手机连接,扫描一次性配对码后,在手机查看指定项目的同一Task/Run、进展、实际文件差异和检查结果;在授予审批权限后,可确认原审批或停止所属任务。离开电脑、关闭桌面窗口不影响后台宿主,手机断线后从原seq补发,撤销设备立即失去访问。执行文件、命令、模型和Git仍由原宿主控制。
首版为PWA伴随端,覆盖观察、原审批、停止和原报告,不在手机上另做一个执行内核。真实手机验收和390px兼容分别记录。完整手机编辑器、原生Android/iOS包、跨公网托管和手机本地执行是后续独立能力。
## 六个切片与门槛
| 切片 | 实现范围 | 验收结果 |
| --- | --- | --- |
| A:配对网关 | 执行宿主继续只监听loopback;用户显式开启独立HTTPS网关;一次性256位配对随机值、短有效期、原宿主/用户/网关身份绑定、串行消费和设备登记 | 错误宿主、过期/重复/并发配对、未知设备与原服务变化拒绝;关闭网关不结束任务;不把原Workbench管理token交给手机 |
| B:设备权限 | 每个设备选择项目范围和观察/审批/停止权限;服务端路由及归属检查;撤销/到期、会话轮换、连接速率限制及原操作审计 | 越项目、换Task/Run、字段扩展、绕过UI直接调用、已撤销设备和旧会话拒绝;设备审计不保存凭据明文 |
| C:手机界面 | 配对、项目/任务列表、原进展、原文件差异、工程检查、预算/unknown状态和报告;触控、软键盘、安全区及前后台恢复 | 在实际Android或iOS浏览器操作;长路径/大量进展/原审批可读,后台切回状态正确;宽窄界面共用原Task/Run和工件 |
| D:审批与停止 | 手机确认同一原审批ID、Task/Run、输入指纹和当前修订;重复请求幂等;明确停止等待所属工具及进程结束 | 桌面与手机同时审批只执行一次;已取消/过期/换Run审批拒绝;断线期间完成的工具不重放;停止后原心跳不再写入 |
| E:断线与撤销 | 从原持久事件seq分页补发,设备仅收到授权项目;重复/乱序处理,原结果SHA核验;撤销立即关闭活动连接 | Wi-Fi断开/切网/刷新/后台/重连均恢复同一结果;无越项目历史;设备撤销时仍在运行的原Task继续由宿主保留 |
| F:独立发行与回顾 | 固定源码/设备契约、全量历史回归、真实手机/桌面、干净安装、保留连接和全部持久数据升级 | 原源码/测试基线和失败保留;手机真实设备信息、步骤/结果/日志有证据;没有真实设备时不能把手机交付标完整通过 |
## 传输和凭据选择
默认关闭远程监听。局域网首版需要手机实际信任的HTTPS证书和固定网关地址;本机生成证书与设备上的信任安装要有明确操作和实测结果。仅访问一个HTTP局域网地址或忽略证书错误,不算PWA与加密传输完成。已有可信域名/证书可作为用户选择的配置,公网或第三方隧道另行配置,不自动公开工作台。
配对随机值仅用于短期单次授权。设备登记与权限修订在宿主事务中保存;持久秘密使用本机加密适配层,手机侧采用安全会话或实际平台可支持的设备密钥。具体认证格式通过实际浏览器验证后固定,不预先宣称硬件密钥或跨平台安全存储已可用。原模型Key始终留在Provider;手机模型状态和用量不包含Key。
## 模块与原契约
拟新增gateway传输/设备登记、共享设备与审批契约、配对管理界面及mobile伴随页面。Runtime、Store、Provider、Task/Run/Tool、工件、计划、累计预算、worker和受控引擎仍复用;新增网关只转发经设备权限和归属核对的固定操作。配对、撤销和设备审计不触发模型或项目命令。
手机不能访问Workbench完整管理API、任意宿主路径、任意命令或全部会话导出。观察权限只读取授权工程;审批或停止需要原身份和当前输入,UI隐藏按钮不是权限边界。每个操作写入设备身份/请求ID/Task/Run/结果,重试不产生第二次副作用。unknown必须显示需核对,不能因断线自动恢复工具。
## 实际验收素材
使用自有Node/React/Python项目、独立端口、数据目录及测试凭据,复用既有失败修复/父审查/原审批/检查/Git样本。设备端保存原步骤和截图,宿主保存原事件/工具行及工件SHA;分别覆盖只读设备、审批设备、撤销、原审批竞争、断线和停止。最终正式升级核对原PID/Provider、全部表每行与全部会话/工件,测试不读取或复制真实模型Key。
每个切片通过相关验收后回顾并本地Git提交,原失败候选不部署。A–E可以先以自有协议/浏览器夹具开发,但F必须增加真实手机结果;需要的设备或证书选择在实际接入前取得用户信息,不能用等待时间替代授权。
## 后续能力顺序
手机完成后,按实际协议接入更多模型,验证工具调用、用量、流式断开和池路由,再做视觉/语音附件、项目范围缓存;最后用独立分支、固定工程评测、多模型复核和可回退候选开发自我迭代。Codex、Claude Code、Cursor、WorkBuddy适配和Purge真实收益均以具体版本/原协议/原工程结果判定,不能以名称或静态研究替代接入。参见[NEXT-STEPS](NEXT-STEPS.md)。
+3 -3
View File
@@ -1,10 +1,10 @@
# Agent Workbench 后续产品与执行开发方案
更新:2026-10-31。本文保留完整产品规划。1.12.1[里程碑审查](MILESTONE-AUDIT.md)与五类确认问题修复、1.13两个受限worker保持;当前正式基线1.14.0已交付固定DeepSeek受控引擎、工具/模型桥、父工程审查/Git及默认关闭的来源锁定提示实验。独立445/445、完整178组UI、三工程13节点及最终归档/便携/保留状态升级通过,原Provider/PID、模型中心与28会话全部历史保持,部署后14页通过。下一阶段按[1.15方案](PLAN-1.15.md)开发Windows桌面与本机加密连接,再按[后续计划](NEXT-STEPS.md)推进手机和其余能力。真实模型工程与Purge推理收益继续按用户选择暂缓;实际实现、验收与部署分别见[能力矩阵](CAPABILITIES.md)及[验证记录](VALIDATION.md)。
更新:2026-10-31。当前正式基线1.14.0、原Provider/PID与全部历史保持。1.15 Windows桌面与本机加密连接的A–E已实现、实际验收并提交;F正在核对独立源码/完整历史UI/最终安装/便携和保留状态部署,未完成前不登记正式交付。下一[1.16手机方案](PLAN-1.16.md)已封存,按[后续计划](NEXT-STEPS.md)逐切片回顾和本地Git。真实模型工程及Purge推理收益继续按用户选择暂缓;具体实现/证据/部署分别见[能力矩阵](CAPABILITIES.md)、[QA-1.15](QA-1.15.md)与[验证记录](VALIDATION.md)。
## 1. 产品目标
1.14的A–F已完成实现、独立验收和正式交付:253份源码输入/6构建/40基准、三工程13节点、23类178组UI及独立445/445通过;最终不可变20261030235440461便携/归档与保留状态升级通过。原PID23172/Provider、全部7表每行及28会话导出保持,具体原结果见[QA-1.14](QA-1.14.md)。[1.15桌面方案](PLAN-1.15.md)记录窗口/宿主分离、本机加密连接、单实例、实际安装/更新/回退六切片,前置已通过,尚未实现。
1.14的A–F已完成实现、独立验收和正式交付:253份源码输入/6构建/40基准、三工程13节点、23类178组UI及独立445/445通过;最终不可变20261030235440461便携/归档与保留状态升级通过。原PID23172/Provider、全部7表每行及28会话导出保持,具体原结果见[QA-1.14](QA-1.14.md)。[1.15桌面方案](PLAN-1.15.md)的A–E已分别实现及实际验收(249f191、72a496a、5354d95、f7a5969、b77783b),F独立发行/正式部署正在执行,后续真实手机配对另属1.16。
建立由我们掌握源码、可接用户选择的大模型、在本地持续完成真实工程任务的AI编程与能力研究工作台。研究Codex、Claude Code、Cursor、DeepSeek Harness、WorkBuddy的有价值机制,用REA取得可追溯证据,将机制独立开发成工作台内置能力;DSH-purge作为明确版本和作用范围的研究及增强策略候选。
@@ -31,7 +31,7 @@ Purge的源码机制不证明模型权重或工程成功率已经提升。增强
现有基础:React工作台、兼容模型连接、Task/Run/Tool及预算、SQLite/工件恢复、工程索引与精确补丁、npm实际检查、单文件/整批审核、Git和报告。1.4.0增加原/新字节快照和逐文件执行账本、整批预检及显式恢复;各版本实测证据单独登记。
正式1.12基础包括Task所属依赖计划与原结果核验、多连接兼容Chat协议模型池、本地格式导入/巡检、带来源核对和有界请求快照的[项目记忆](PROJECT-MEMORY.md)、Task检查点、本地Skills和受控stdio MCP。1.13源码已实现两个受限worker、独立Git工作区、累计预算/停止恢复、当前成果核对和父工程文本审查合并,最终发行以[QA-1.13](QA-1.13.md)为准。动态REA/外部引擎注册、任意规模调度、OS沙箱、桌面壳和手机配对仍待开发;已有不同阶段的Node/React/Python工程样本,整体自主交付能力按真实任务评估。Python专用检查和混合范围选择的来源见[Python契约](PYTHON-CHECKS.md),真实React类型/构建/交互/Git样本见[React基准](../benchmarks/react-delivery-v1/README.md)。任意自动合并与归档继续开发,现有批次支持文本创建/修改,不提供跨文件原子事务。当前模型协议仍使用stream:false;跨协议与真实模型池工程验收另行记录。
正式1.12基础包括Task所属依赖计划与原结果核验、多连接兼容Chat协议模型池、本地格式导入/巡检、带来源核对和有界请求快照的[项目记忆](PROJECT-MEMORY.md)、Task检查点、本地Skills和受控stdio MCP。1.13源码已实现两个受限worker、独立Git工作区、累计预算/停止恢复、当前成果核对和父工程文本审查合并,最终发行以[QA-1.13](QA-1.13.md)为准。1.14受控DeepSeek工程引擎已交付;1.15桌面宿主/窗口、本机加密和安装A–E已通过,正式发行看QA-1.15。动态REA/任意外部引擎注册、任意规模调度、OS沙箱及真实手机配对仍待开发;已有不同阶段的Node/React/Python工程样本,整体自主交付能力按真实任务评估。Python专用检查和混合范围选择的来源见[Python契约](PYTHON-CHECKS.md),真实React类型/构建/交互/Git样本见[React基准](../benchmarks/react-delivery-v1/README.md)。任意自动合并与归档继续开发,现有批次支持文本创建/修改,不提供跨文件原子事务。当前模型协议仍使用stream:false;跨协议与真实模型池工程验收另行记录。
工作台源码已建立本地Git,基线提交为`bac87da`,标签`v1.0.0-baseline`,继续按可审查小步提交。用户工程已有内置审查提交/证据报告、独立工作区及1.13父子文本成果审查合并;任意自动合并、归档和总体自主工程完成率继续实施与验收。
+2 -2
View File
@@ -52,6 +52,6 @@ C#旧compiler不支持deterministic选项,使用源码/图标/编译器SHA绑
原环境根误用、精确seq误判、父审查漏点“审查变更集”、隐藏窗口截图等待绘制、Playwright下载事件、截图后的弹窗遮挡、窄屏自动收起项目栏均保存原失败。修正到实际SDK根、原事件补发、正常界面动作、显示窗口后捕获和原生DownloadItem,不使用force点击或放宽工程/预算/审批断言。[67份原证据](desktop-projects-e-evidence.json)包含失败/通过及PNG/DOM/日志;完整原fixture/工件仍保存在所引QA目录。原UI9组之后,仅共享QA helper增加正常弹窗关闭/项目栏展开,[原/新SHA差异](desktop-projects-e-ui-helper-delta.json)独立登记;产品/构建未改变,F以最终1.15重新验收这个helper。D本地提交f7a5969;E仍是开发版1.14.0,F不可跳过。
## F:独立发行与正式部署(待执行)
## F:独立发行与正式部署(进行中)
完成E并提交后,固定1.15.0输入,核对全部构建目录、完整源码和历史UI,再验收源码/归档/便携/实际安装及保留原宿主升级。按照[执行方案](PLAN-1.15.md)完成F后才将本阶段计为完整交付。
E已提交b77783b。已统一1.15.0元数据/独立证据路径、离线随包Electron和实际源码构建输入;源码重建资产加入宿主身份。首个构建误用系统Node16,原日志保留;固定Node24.22.1后构建及17相关检查通过。当前固定1.15.0输入,核对全部构建目录、完整源码和历史UI,再验收源码/归档/便携/实际安装及保留原宿主升级。按照[执行方案](PLAN-1.15.md)完成F后才将本阶段计为完整交付。
+5
View File
@@ -43,3 +43,8 @@
新适配脚本暴露了几类原假设:父成果按钮只导航到改动区,还需实际审查动作;隐藏Electron窗口截图可能不绘制;Playwright网页下载事件不适合作为本机下载结论;宽窄切换后,弹窗和项目栏状态仍需正常操作。原失败均保存,补正常按钮/关闭/展开及实际DownloadItem,避免把自动化穿透遮罩当真实用户流程。原生DownloadItem落盘和原SHA证明下载过程,Windows保存对话框仍未测。
本切片继续用自有HTTP模型响应算法和in-process宿主;工程命令、检查、文件审查、Git和原生进程实际运行,但真实模型推理与整宿主退出需要分别判定。当前实际停止/崩溃/重启由C/D覆盖。原9组UI之后的helper小改动单独保存原/新SHA,F最终版本复验;历史报告不重写。下一F同步全部版本/独立证据路径、离线源码含Electron运行依赖及全部构建范围,再完整回归、实际安装和保留状态正式部署。
## F发行前回顾
A–E原证据和五个提交保持。F把离线源码Electron运行、重建客户端资产身份和全部dist构建范围纳入验收;完整回归/新归档/实际安装/状态保留部署继续执行。首个npm构建由系统旧Node16抢占,固定子进程PATH为实际Node24.22.1后通过,原失败未覆盖。当前17宿主/窗口/安装/旧升级专项通过,不代表最终整轮已通过。
+4
View File
@@ -1,5 +1,9 @@
# Agent Workbench 验收记录
## 1.15 · Windows桌面与本机加密(A–E通过,F验收中)
宿主身份/单实例、DPAPI选择保存、实际Electron窗口、安装/回退/卸载与分页工件巡检、桌面三工程16节点和原生UI9组已通过;五切片分别提交。F正进行1.15最终源码/历史UI/安装/便携及保留状态正式部署,当前正式仍1.14.0/PID23172;不将局部结果计为完整发行。原失败与具体范围见[QA-1.15](QA-1.15.md)、[回顾](RETROSPECTIVE-1.15.md)与[下一手机方案](PLAN-1.16.md)。
## 1.14 · 受控引擎、默认关闭提示实验与正式交付
固定DeepSeek Harness的受限工具/模型桥、所属独立工程、原prompt/审批/预算/结果、停止核对和父工程审查/检查/Git已交付;两版本Purge来源锁定的提示实验默认关闭。新Node/React/Python三工程13节点、父8/8、9/9、10/10及构建/编译/Git、React真实Edge两视口各8项断言通过。独立源码445/445、253输入/6构建/40基准、23类178组完整UI及7阶段原日志通过;最终不可变20261030235440461便携11427/源码9748文件归档与实际随包受控四节点通过。正式4318保留PID23172/Provider升级1.14.0,2连接/1池/6原请求摘要、16项目/28全部会话/2551记录/4568事件及全部7表每行/全部导出保持,部署后14页通过,9229关闭。升级后QA误判原Node路径及SQLite行原型的原失败分别保留,续验没有重放升级;本轮真实模型0请求。见[1.14验收](QA-1.14.md)、[回顾](RETROSPECTIVE-1.14.md)、[发行身份](controlled-engine-release-identity.json)与[下一桌面计划](PLAN-1.15.md)。
+94 -94
View File
@@ -1,95 +1,95 @@
{
"success": true,
"root": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525",
"url": "http://127.0.0.1:62117",
"stage": "plaintext leak scan over actual database, WAL, exports and host logs",
"originalHost": {
"url": "http://127.0.0.1:62117",
"identity": {
"schema": 1,
"protocol": "agent-workbench-host-v1",
"app": "agent-workbench",
"version": "1.14.0",
"instanceId": "ad80c3c8-5484-443d-b142-1d763e49350f",
"pid": 28724,
"createdAt": "2026-10-31T01:21:14.095Z",
"userSid": "S-1-5-21-3906741337-925321518-3767092861-1001",
"ownerId": "41b7bfe9-e7f5-4051-a5bc-e371575da94a",
"executable": {
"path": "D:\\Ai-project\\REA逆向\\tools\\node-v24.22.1-win-x64\\node.exe",
"bytes": 93915464,
"sha256": "d702daa815eabbff763c7d3808e235cf031cf2ca70211b56847bc72f51e240e5"
},
"appRoot": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525\\app",
"runtimeRoot": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525\\app",
"runtimeIdentity": "e53c216c4180c7b2960028f2e9cb1763fc5510efe0ea088e5370450e76a621d8",
"dataRoot": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525\\data",
"port": 62117
},
"reused": false,
"logs": {
"stdoutPath": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525\\data\\logs\\desktop-host-78659811-d287-4dd2-bee4-655a36c0c328-stdout.log",
"stderrPath": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525\\data\\logs\\desktop-host-78659811-d287-4dd2-bee4-655a36c0c328-stderr.log"
}
},
"freshHost": {
"url": "http://127.0.0.1:62117",
"identity": {
"schema": 1,
"protocol": "agent-workbench-host-v1",
"app": "agent-workbench",
"version": "1.14.0",
"instanceId": "ae5ee3e1-584d-4969-bbe3-7483382035cb",
"pid": 49924,
"createdAt": "2026-10-31T01:21:23.749Z",
"userSid": "S-1-5-21-3906741337-925321518-3767092861-1001",
"ownerId": "5d0fb73d-77b2-450a-9a07-b089a3e6a4c6",
"executable": {
"path": "D:\\Ai-project\\REA逆向\\tools\\node-v24.22.1-win-x64\\node.exe",
"bytes": 93915464,
"sha256": "d702daa815eabbff763c7d3808e235cf031cf2ca70211b56847bc72f51e240e5"
},
"appRoot": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525\\app",
"runtimeRoot": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525\\app",
"runtimeIdentity": "e53c216c4180c7b2960028f2e9cb1763fc5510efe0ea088e5370450e76a621d8",
"dataRoot": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525\\data",
"port": 62117
},
"reused": false,
"logs": {
"stdoutPath": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525\\data\\logs\\desktop-host-747ea62f-e2db-46e7-a878-d48d3533ab64-stdout.log",
"stderrPath": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525\\data\\logs\\desktop-host-747ea62f-e2db-46e7-a878-d48d3533ab64-stderr.log"
}
},
"groups": [
"real connection form and explicit encrypted save",
"real legacy credential save without exposing an existing key",
"real desktop and 390px saved-state controls and layout",
"actual host exit and fresh process restores encrypted revisions",
"fresh host actually uses both DPAPI-restored keys with owned HTTP responses",
"disable, delete and forget through real controls",
"plaintext leak scan over actual database, WAL, exports and host logs"
],
"requests": [
{
"path": "/v1/chat/completions",
"authorizedOwnedKey": true
},
{
"path": "/v1/chat/completions",
"authorizedOwnedKey": true
},
{
"path": "/v1/chat/completions",
"authorizedOwnedKey": true
}
],
"images": [
"D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525\\credentials-new-connection.png",
"D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525\\credentials-1440.png",
"D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525\\credentials-390.png"
],
"pageErrors": [],
"detail": null,
"scope": "Owned CurrentUser DPAPI and local HTTP fixture; no external model request or existing credential is used."
{
"success": true,
"root": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525",
"url": "http://127.0.0.1:62117",
"stage": "plaintext leak scan over actual database, WAL, exports and host logs",
"originalHost": {
"url": "http://127.0.0.1:62117",
"identity": {
"schema": 1,
"protocol": "agent-workbench-host-v1",
"app": "agent-workbench",
"version": "1.14.0",
"instanceId": "ad80c3c8-5484-443d-b142-1d763e49350f",
"pid": 28724,
"createdAt": "2026-10-31T01:21:14.095Z",
"userSid": "S-1-5-21-3906741337-925321518-3767092861-1001",
"ownerId": "41b7bfe9-e7f5-4051-a5bc-e371575da94a",
"executable": {
"path": "D:\\Ai-project\\REA逆向\\tools\\node-v24.22.1-win-x64\\node.exe",
"bytes": 93915464,
"sha256": "d702daa815eabbff763c7d3808e235cf031cf2ca70211b56847bc72f51e240e5"
},
"appRoot": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525\\app",
"runtimeRoot": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525\\app",
"runtimeIdentity": "e53c216c4180c7b2960028f2e9cb1763fc5510efe0ea088e5370450e76a621d8",
"dataRoot": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525\\data",
"port": 62117
},
"reused": false,
"logs": {
"stdoutPath": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525\\data\\logs\\desktop-host-78659811-d287-4dd2-bee4-655a36c0c328-stdout.log",
"stderrPath": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525\\data\\logs\\desktop-host-78659811-d287-4dd2-bee4-655a36c0c328-stderr.log"
}
},
"freshHost": {
"url": "http://127.0.0.1:62117",
"identity": {
"schema": 1,
"protocol": "agent-workbench-host-v1",
"app": "agent-workbench",
"version": "1.14.0",
"instanceId": "ae5ee3e1-584d-4969-bbe3-7483382035cb",
"pid": 49924,
"createdAt": "2026-10-31T01:21:23.749Z",
"userSid": "S-1-5-21-3906741337-925321518-3767092861-1001",
"ownerId": "5d0fb73d-77b2-450a-9a07-b089a3e6a4c6",
"executable": {
"path": "D:\\Ai-project\\REA逆向\\tools\\node-v24.22.1-win-x64\\node.exe",
"bytes": 93915464,
"sha256": "d702daa815eabbff763c7d3808e235cf031cf2ca70211b56847bc72f51e240e5"
},
"appRoot": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525\\app",
"runtimeRoot": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525\\app",
"runtimeIdentity": "e53c216c4180c7b2960028f2e9cb1763fc5510efe0ea088e5370450e76a621d8",
"dataRoot": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525\\data",
"port": 62117
},
"reused": false,
"logs": {
"stdoutPath": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525\\data\\logs\\desktop-host-747ea62f-e2db-46e7-a878-d48d3533ab64-stdout.log",
"stderrPath": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525\\data\\logs\\desktop-host-747ea62f-e2db-46e7-a878-d48d3533ab64-stderr.log"
}
},
"groups": [
"real connection form and explicit encrypted save",
"real legacy credential save without exposing an existing key",
"real desktop and 390px saved-state controls and layout",
"actual host exit and fresh process restores encrypted revisions",
"fresh host actually uses both DPAPI-restored keys with owned HTTP responses",
"disable, delete and forget through real controls",
"plaintext leak scan over actual database, WAL, exports and host logs"
],
"requests": [
{
"path": "/v1/chat/completions",
"authorizedOwnedKey": true
},
{
"path": "/v1/chat/completions",
"authorizedOwnedKey": true
},
{
"path": "/v1/chat/completions",
"authorizedOwnedKey": true
}
],
"images": [
"D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525\\credentials-new-connection.png",
"D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525\\credentials-1440.png",
"D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525\\credentials-390.png"
],
"pageErrors": [],
"detail": null,
"scope": "Owned CurrentUser DPAPI and local HTTP fixture; no external model request or existing credential is used."
}
+7 -7
View File
@@ -1,7 +1,7 @@
PASS real connection form and explicit encrypted save
PASS real legacy credential save without exposing an existing key
PASS real desktop and 390px saved-state controls and layout
PASS actual host exit and fresh process restores encrypted revisions
PASS fresh host actually uses both DPAPI-restored keys with owned HTTP responses
PASS disable, delete and forget through real controls
PASS plaintext leak scan over actual database, WAL, exports and host logs
PASS real connection form and explicit encrypted save
PASS real legacy credential save without exposing an existing key
PASS real desktop and 390px saved-state controls and layout
PASS actual host exit and fresh process restores encrypted revisions
PASS fresh host actually uses both DPAPI-restored keys with owned HTTP responses
PASS disable, delete and forget through real controls
PASS plaintext leak scan over actual database, WAL, exports and host logs
+36 -36
View File
@@ -1,37 +1,37 @@
{
"success": true,
"root": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4cfcf5da-eaaf-472e-b6c0-99266b1e9a3c",
"url": "http://127.0.0.1:62578",
"stage": "plaintext leak scan over actual database, WAL, exports and host logs",
"groups": [
"real connection form and explicit encrypted save",
"real legacy credential save without exposing an existing key",
"real desktop and 390px saved-state controls and layout",
"actual host exit and fresh process restores encrypted revisions",
"fresh host actually uses both DPAPI-restored keys with owned HTTP responses",
"disable, delete and forget through real controls",
"plaintext leak scan over actual database, WAL, exports and host logs"
],
"requests": [
{
"path": "/v1/chat/completions",
"authorizedOwnedKey": true
},
{
"path": "/v1/chat/completions",
"authorizedOwnedKey": true
},
{
"path": "/v1/chat/completions",
"authorizedOwnedKey": true
}
],
"images": [
"D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4cfcf5da-eaaf-472e-b6c0-99266b1e9a3c\\credentials-new-connection.png",
"D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4cfcf5da-eaaf-472e-b6c0-99266b1e9a3c\\credentials-1440.png",
"D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4cfcf5da-eaaf-472e-b6c0-99266b1e9a3c\\credentials-390.png"
],
"pageErrors": [],
"detail": null,
"scope": "Owned CurrentUser DPAPI and local HTTP fixture; no external model request or existing credential is used."
{
"success": true,
"root": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4cfcf5da-eaaf-472e-b6c0-99266b1e9a3c",
"url": "http://127.0.0.1:62578",
"stage": "plaintext leak scan over actual database, WAL, exports and host logs",
"groups": [
"real connection form and explicit encrypted save",
"real legacy credential save without exposing an existing key",
"real desktop and 390px saved-state controls and layout",
"actual host exit and fresh process restores encrypted revisions",
"fresh host actually uses both DPAPI-restored keys with owned HTTP responses",
"disable, delete and forget through real controls",
"plaintext leak scan over actual database, WAL, exports and host logs"
],
"requests": [
{
"path": "/v1/chat/completions",
"authorizedOwnedKey": true
},
{
"path": "/v1/chat/completions",
"authorizedOwnedKey": true
},
{
"path": "/v1/chat/completions",
"authorizedOwnedKey": true
}
],
"images": [
"D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4cfcf5da-eaaf-472e-b6c0-99266b1e9a3c\\credentials-new-connection.png",
"D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4cfcf5da-eaaf-472e-b6c0-99266b1e9a3c\\credentials-1440.png",
"D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4cfcf5da-eaaf-472e-b6c0-99266b1e9a3c\\credentials-390.png"
],
"pageErrors": [],
"detail": null,
"scope": "Owned CurrentUser DPAPI and local HTTP fixture; no external model request or existing credential is used."
}
+7 -7
View File
@@ -1,7 +1,7 @@
PASS real connection form and explicit encrypted save
PASS real legacy credential save without exposing an existing key
PASS real desktop and 390px saved-state controls and layout
PASS actual host exit and fresh process restores encrypted revisions
PASS fresh host actually uses both DPAPI-restored keys with owned HTTP responses
PASS disable, delete and forget through real controls
PASS plaintext leak scan over actual database, WAL, exports and host logs
PASS real connection form and explicit encrypted save
PASS real legacy credential save without exposing an existing key
PASS real desktop and 390px saved-state controls and layout
PASS actual host exit and fresh process restores encrypted revisions
PASS fresh host actually uses both DPAPI-restored keys with owned HTTP responses
PASS disable, delete and forget through real controls
PASS plaintext leak scan over actual database, WAL, exports and host logs
+136 -136
View File
@@ -1,136 +1,136 @@
TAP version 13
# Subtest: actual Windows CurrentUser DPAPI roundtrips owned UTF-8 and rejects damaged bytes and wrong entropy
ok 1 - actual Windows CurrentUser DPAPI roundtrips owned UTF-8 and rejects damaged bytes and wrong entropy
---
duration_ms: 1703.112
type: 'test'
...
# Subtest: only explicitly supplied keys persist; new provider restores original revisions and public views contain no key
ok 2 - only explicitly supplied keys persist; new provider restores original revisions and public views contain no key
---
duration_ms: 2823.1817
type: 'test'
...
# Subtest: registry and cipher commit failure restores original routing, keys, block and every database row
ok 3 - registry and cipher commit failure restores original routing, keys, block and every database row
---
duration_ms: 1872.4547
type: 'test'
...
# Subtest: a two-key import with a protection failure rolls back all registry and cipher entries and retains its preview
ok 4 - a two-key import with a protection failure rolls back all registry and cipher entries and retains its preview
---
duration_ms: 1892.0773
type: 'test'
...
# Subtest: connection revision changes retain the original saved block and report a source mismatch on restart
ok 5 - connection revision changes retain the original saved block and report a source mismatch on restart
---
duration_ms: 2825.6882
type: 'test'
...
# Subtest: disable skips real restart loading, explicit enable restores, delete preserves memory and forget removes both
ok 6 - disable skips real restart loading, explicit enable restores, delete preserves memory and forget removes both
---
duration_ms: 2609.0633
type: 'test'
...
# Subtest: corrupt saved block and another SID do not load; failures are visible without returning ciphertext or plaintext
ok 7 - corrupt saved block and another SID do not load; failures are visible without returning ciphertext or plaintext
---
duration_ms: 2682.8404
type: 'test'
...
# Subtest: legacy persistence requires a new explicit key and binds config and credential revisions; stale requests cannot save
ok 8 - legacy persistence requires a new explicit key and binds config and credential revisions; stale requests cannot save
---
duration_ms: 2361.2135
type: 'test'
...
# Subtest: owned plaintext keys never reach SQLite pages, WAL, config or public execution exports
ok 9 - owned plaintext keys never reach SQLite pages, WAL, config or public execution exports
---
duration_ms: 1256.7214
type: 'test'
...
# Subtest: model registry keeps original credentials/endpoint intact, rejects stale edits, and forgets only volatile keys on restart
ok 10 - model registry keeps original credentials/endpoint intact, rejects stale edits, and forgets only volatile keys on restart
---
duration_ms: 914.4979
type: 'test'
...
# Subtest: config import previews exact revisions, keeps secrets volatile, syncs only when enabled and preserves manual edits on conflicts
ok 11 - config import previews exact revisions, keeps secrets volatile, syncs only when enabled and preserves manual edits on conflicts
---
duration_ms: 753.8609
type: 'test'
...
# Subtest: real HTTP pool fails over 429 to another model and never repeats completed tools; attempts seal actual request bytes and Run identity
ok 12 - real HTTP pool fails over 429 to another model and never repeats completed tools; attempts seal actual request bytes and Run identity
---
duration_ms: 2729.1166
type: 'test'
...
# Subtest: pool rotation and weights select members on new Runs; disabled and missing credentials are excluded without requests
ok 13 - pool rotation and weights select members on new Runs; disabled and missing credentials are excluded without requests
---
duration_ms: 3857.7792
type: 'test'
...
# Subtest: 401 blocks the credential revision; updating the key restores availability without leaking error echoes
ok 14 - 401 blocks the credential revision; updating the key restores availability without leaking error echoes
---
duration_ms: 2080.9134
type: 'test'
...
# Subtest: queued request cancellation and in-flight disable release only owned leases and do not dispatch another HTTP request
ok 15 - queued request cancellation and in-flight disable release only owned leases and do not dispatch another HTTP request
---
duration_ms: 958.8689
type: 'test'
...
# Subtest: compiled model-center HTTP API persists public profiles, blocks forged origins, and local operations never invoke a model
ok 16 - compiled model-center HTTP API persists public profiles, blocks forged origins, and local operations never invoke a model
---
duration_ms: 2084.6897
type: 'test'
...
# Subtest: model registry disk failures restore routing, pool profiles and credential references without partial updates
ok 17 - model registry disk failures restore routing, pool profiles and credential references without partial updates
---
duration_ms: 957.6324
type: 'test'
...
# Subtest: model selection and saved configuration commit together and failures preserve the original mode and endpoint
ok 18 - model selection and saved configuration commit together and failures preserve the original mode and endpoint
---
duration_ms: 711.9416
type: 'test'
...
# Subtest: configuration imports roll back the entire batch and retain one usable preview after validation or commit failure
ok 19 - configuration imports roll back the entire batch and retain one usable preview after validation or commit failure
---
duration_ms: 726.4839
type: 'test'
...
# Subtest: hot upgrade refreshes retained model and importer methods while preserving credential and routing objects
ok 20 - hot upgrade refreshes retained model and importer methods while preserving credential and routing objects
---
duration_ms: 751.8999
type: 'test'
...
# Subtest: hot upgrade refuses active probes and restores retained model methods if installation fails
ok 21 - hot upgrade refuses active probes and restores retained model methods if installation fails
---
duration_ms: 728.8236
type: 'test'
...
1..21
# tests 21
# suites 0
# pass 21
# fail 0
# cancelled 0
# skipped 0
# todo 0
# duration_ms 20118.8961
TAP version 13
# Subtest: actual Windows CurrentUser DPAPI roundtrips owned UTF-8 and rejects damaged bytes and wrong entropy
ok 1 - actual Windows CurrentUser DPAPI roundtrips owned UTF-8 and rejects damaged bytes and wrong entropy
---
duration_ms: 1703.112
type: 'test'
...
# Subtest: only explicitly supplied keys persist; new provider restores original revisions and public views contain no key
ok 2 - only explicitly supplied keys persist; new provider restores original revisions and public views contain no key
---
duration_ms: 2823.1817
type: 'test'
...
# Subtest: registry and cipher commit failure restores original routing, keys, block and every database row
ok 3 - registry and cipher commit failure restores original routing, keys, block and every database row
---
duration_ms: 1872.4547
type: 'test'
...
# Subtest: a two-key import with a protection failure rolls back all registry and cipher entries and retains its preview
ok 4 - a two-key import with a protection failure rolls back all registry and cipher entries and retains its preview
---
duration_ms: 1892.0773
type: 'test'
...
# Subtest: connection revision changes retain the original saved block and report a source mismatch on restart
ok 5 - connection revision changes retain the original saved block and report a source mismatch on restart
---
duration_ms: 2825.6882
type: 'test'
...
# Subtest: disable skips real restart loading, explicit enable restores, delete preserves memory and forget removes both
ok 6 - disable skips real restart loading, explicit enable restores, delete preserves memory and forget removes both
---
duration_ms: 2609.0633
type: 'test'
...
# Subtest: corrupt saved block and another SID do not load; failures are visible without returning ciphertext or plaintext
ok 7 - corrupt saved block and another SID do not load; failures are visible without returning ciphertext or plaintext
---
duration_ms: 2682.8404
type: 'test'
...
# Subtest: legacy persistence requires a new explicit key and binds config and credential revisions; stale requests cannot save
ok 8 - legacy persistence requires a new explicit key and binds config and credential revisions; stale requests cannot save
---
duration_ms: 2361.2135
type: 'test'
...
# Subtest: owned plaintext keys never reach SQLite pages, WAL, config or public execution exports
ok 9 - owned plaintext keys never reach SQLite pages, WAL, config or public execution exports
---
duration_ms: 1256.7214
type: 'test'
...
# Subtest: model registry keeps original credentials/endpoint intact, rejects stale edits, and forgets only volatile keys on restart
ok 10 - model registry keeps original credentials/endpoint intact, rejects stale edits, and forgets only volatile keys on restart
---
duration_ms: 914.4979
type: 'test'
...
# Subtest: config import previews exact revisions, keeps secrets volatile, syncs only when enabled and preserves manual edits on conflicts
ok 11 - config import previews exact revisions, keeps secrets volatile, syncs only when enabled and preserves manual edits on conflicts
---
duration_ms: 753.8609
type: 'test'
...
# Subtest: real HTTP pool fails over 429 to another model and never repeats completed tools; attempts seal actual request bytes and Run identity
ok 12 - real HTTP pool fails over 429 to another model and never repeats completed tools; attempts seal actual request bytes and Run identity
---
duration_ms: 2729.1166
type: 'test'
...
# Subtest: pool rotation and weights select members on new Runs; disabled and missing credentials are excluded without requests
ok 13 - pool rotation and weights select members on new Runs; disabled and missing credentials are excluded without requests
---
duration_ms: 3857.7792
type: 'test'
...
# Subtest: 401 blocks the credential revision; updating the key restores availability without leaking error echoes
ok 14 - 401 blocks the credential revision; updating the key restores availability without leaking error echoes
---
duration_ms: 2080.9134
type: 'test'
...
# Subtest: queued request cancellation and in-flight disable release only owned leases and do not dispatch another HTTP request
ok 15 - queued request cancellation and in-flight disable release only owned leases and do not dispatch another HTTP request
---
duration_ms: 958.8689
type: 'test'
...
# Subtest: compiled model-center HTTP API persists public profiles, blocks forged origins, and local operations never invoke a model
ok 16 - compiled model-center HTTP API persists public profiles, blocks forged origins, and local operations never invoke a model
---
duration_ms: 2084.6897
type: 'test'
...
# Subtest: model registry disk failures restore routing, pool profiles and credential references without partial updates
ok 17 - model registry disk failures restore routing, pool profiles and credential references without partial updates
---
duration_ms: 957.6324
type: 'test'
...
# Subtest: model selection and saved configuration commit together and failures preserve the original mode and endpoint
ok 18 - model selection and saved configuration commit together and failures preserve the original mode and endpoint
---
duration_ms: 711.9416
type: 'test'
...
# Subtest: configuration imports roll back the entire batch and retain one usable preview after validation or commit failure
ok 19 - configuration imports roll back the entire batch and retain one usable preview after validation or commit failure
---
duration_ms: 726.4839
type: 'test'
...
# Subtest: hot upgrade refreshes retained model and importer methods while preserving credential and routing objects
ok 20 - hot upgrade refreshes retained model and importer methods while preserving credential and routing objects
---
duration_ms: 751.8999
type: 'test'
...
# Subtest: hot upgrade refuses active probes and restores retained model methods if installation fails
ok 21 - hot upgrade refuses active probes and restores retained model methods if installation fails
---
duration_ms: 728.8236
type: 'test'
...
1..21
# tests 21
# suites 0
# pass 21
# fail 0
# cancelled 0
# skipped 0
# todo 0
# duration_ms 20118.8961
+304 -304
View File
@@ -1,304 +1,304 @@
TAP version 13
# Subtest: host challenge proof rejects a stale challenge, changed identity and changed signature
ok 1 - host challenge proof rejects a stale challenge, changed identity and changed signature
---
duration_ms: 1.1348
type: 'test'
...
# Subtest: stop confirmation binds the original instance, creation and loaded runtime
ok 2 - stop confirmation binds the original instance, creation and loaded runtime
---
duration_ms: 0.1247
type: 'test'
...
# Subtest: state paths reject escaping, linked ancestors, ambiguous hardlinks and damaged JSON
ok 3 - state paths reject escaping, linked ancestors, ambiguous hardlinks and damaged JSON
---
duration_ms: 1133.2191
type: 'test'
...
# Subtest: failed registration publishes no initialized memory and can complete after its actual inputs are supplied
ok 4 - failed registration publishes no initialized memory and can complete after its actual inputs are supplied
---
duration_ms: 2526.6054
type: 'test'
...
# Subtest: host shutdown cancels an un-signalled HTTP model request and refuses later requests until recovered
ok 5 - host shutdown cancels an un-signalled HTTP model request and refuses later requests until recovered
---
duration_ms: 1.9739
type: 'test'
...
# Subtest: host shutdown cancels model-pool acquisition and releases queue ownership without inventing attempts
ok 6 - host shutdown cancels model-pool acquisition and releases queue ownership without inventing attempts
---
duration_ms: 818.2146
type: 'test'
...
# Subtest: a question about the previous workbench error receives local error context and current change states
ok 7 - a question about the previous workbench error receives local error context and current change states
---
duration_ms: 2213.6377
type: 'test'
...
# Subtest: a transient model failure after an approved command retries only the model request
ok 8 - a transient model failure after an approved command retries only the model request
---
duration_ms: 2266.197
type: 'test'
...
# Subtest: exhausted model retries preserve tool results, and explicit resume adds no new user turn or replay
ok 9 - exhausted model retries preserve tool results, and explicit resume adds no new user turn or replay
---
duration_ms: 3657.3963
type: 'test'
...
# Subtest: stopping during model backoff cancels further requests and clears the running request state
ok 10 - stopping during model backoff cancels further requests and clears the running request state
---
duration_ms: 1798.9338
type: 'test'
...
# Subtest: HTTP authentication errors are classified, redacted and never automatically retried
ok 11 - HTTP authentication errors are classified, redacted and never automatically retried
---
duration_ms: 21.9978
type: 'test'
...
# Subtest: temporary HTTP service errors retry and recover with recorded attempt counts
ok 12 - temporary HTTP service errors retry and recover with recorded attempt counts
---
duration_ms: 8.2405
type: 'test'
...
# Subtest: invalid and oversized protocol responses fail once with a useful response-format diagnosis
ok 13 - invalid and oversized protocol responses fail once with a useful response-format diagnosis
---
duration_ms: 4.0449
type: 'test'
...
# Subtest: HTTP rate limits expose their source and obey the service retry interval before recovery
ok 14 - HTTP rate limits expose their source and obey the service retry interval before recovery
---
duration_ms: 1010.8285
type: 'test'
...
# Subtest: Windows commands preserve Chinese through a PowerShell pipeline into Node
ok 15 - Windows commands preserve Chinese through a PowerShell pipeline into Node
---
duration_ms: 1003.6617
type: 'test'
...
# Subtest: Windows shell syntax errors stay readable and native command exit codes are retained
ok 16 - Windows shell syntax errors stay readable and native command exit codes are retained
---
duration_ms: 971.0757
type: 'test'
...
# Subtest: review and completed states do not turn a generic continue click into a new model request
ok 17 - review and completed states do not turn a generic continue click into a new model request
---
duration_ms: 852.3601
type: 'test'
...
# Subtest: old session data receives a truthful historical connection diagnosis without rewriting history
ok 18 - old session data receives a truthful historical connection diagnosis without rewriting history
---
duration_ms: 1178.3301
type: 'test'
...
# Subtest: model registry keeps original credentials/endpoint intact, rejects stale edits, and forgets only volatile keys on restart
ok 19 - model registry keeps original credentials/endpoint intact, rejects stale edits, and forgets only volatile keys on restart
---
duration_ms: 1356.0532
type: 'test'
...
# Subtest: config import previews exact revisions, keeps secrets volatile, syncs only when enabled and preserves manual edits on conflicts
ok 20 - config import previews exact revisions, keeps secrets volatile, syncs only when enabled and preserves manual edits on conflicts
---
duration_ms: 1132.7334
type: 'test'
...
# Subtest: real HTTP pool fails over 429 to another model and never repeats completed tools; attempts seal actual request bytes and Run identity
ok 21 - real HTTP pool fails over 429 to another model and never repeats completed tools; attempts seal actual request bytes and Run identity
---
duration_ms: 3159.0983
type: 'test'
...
# Subtest: pool rotation and weights select members on new Runs; disabled and missing credentials are excluded without requests
ok 22 - pool rotation and weights select members on new Runs; disabled and missing credentials are excluded without requests
---
duration_ms: 4730.0215
type: 'test'
...
# Subtest: 401 blocks the credential revision; updating the key restores availability without leaking error echoes
ok 23 - 401 blocks the credential revision; updating the key restores availability without leaking error echoes
---
duration_ms: 2035.7928
type: 'test'
...
# Subtest: queued request cancellation and in-flight disable release only owned leases and do not dispatch another HTTP request
ok 24 - queued request cancellation and in-flight disable release only owned leases and do not dispatch another HTTP request
---
duration_ms: 1053.5497
type: 'test'
...
# Subtest: compiled model-center HTTP API persists public profiles, blocks forged origins, and local operations never invoke a model
ok 25 - compiled model-center HTTP API persists public profiles, blocks forged origins, and local operations never invoke a model
---
duration_ms: 2488.7937
type: 'test'
...
# Subtest: model registry disk failures restore routing, pool profiles and credential references without partial updates
ok 26 - model registry disk failures restore routing, pool profiles and credential references without partial updates
---
duration_ms: 924.9926
type: 'test'
...
# Subtest: model selection and saved configuration commit together and failures preserve the original mode and endpoint
ok 27 - model selection and saved configuration commit together and failures preserve the original mode and endpoint
---
duration_ms: 896.6963
type: 'test'
...
# Subtest: configuration imports roll back the entire batch and retain one usable preview after validation or commit failure
ok 28 - configuration imports roll back the entire batch and retain one usable preview after validation or commit failure
---
duration_ms: 903.5305
type: 'test'
...
# Subtest: a stop from the event loop interrupts immediate model replies before later tools
ok 29 - a stop from the event loop interrupts immediate model replies before later tools
---
duration_ms: 1924.7977
type: 'test'
...
# Subtest: a stop after the first tool preserves a complete interrupted multi-call protocol group
ok 30 - a stop after the first tool preserves a complete interrupted multi-call protocol group
---
duration_ms: 2188.5101
type: 'test'
...
# Subtest: the compatible HTTP loop preserves call IDs, validates errors and waits before executing commands
ok 31 - the compatible HTTP loop preserves call IDs, validates errors and waits before executing commands
---
duration_ms: 5677.1928
type: 'test'
...
# Subtest: denial prevents command execution and produces a paired denied result
ok 32 - denial prevents command execution and produces a paired denied result
---
duration_ms: 1786.9452
type: 'test'
...
# Subtest: stopping during approval interrupts pending calls without executing or replaying them
ok 33 - stopping during approval interrupts pending calls without executing or replaying them
---
duration_ms: 1887.1945
type: 'test'
...
# Subtest: stopping a running fixture command terminates its own process tree and retains uncertain state
ok 34 - stopping a running fixture command terminates its own process tree and retains uncertain state
---
duration_ms: 1887.357
type: 'test'
...
# Subtest: command output decodes split UTF-8 and excludes the workbench key from child environment
ok 35 - command output decodes split UTF-8 and excludes the workbench key from child environment
---
duration_ms: 1547.6495
type: 'test'
...
# Subtest: model-list testing uses a tentative key without changing the saved connection and redacts service errors
ok 36 - model-list testing uses a tentative key without changing the saved connection and redacts service errors
---
duration_ms: 6.3458
type: 'test'
...
# Subtest: truncated or repeated model calls fail before execution; tool input validation rejects unknown capabilities
ok 37 - truncated or repeated model calls fail before execution; tool input validation rejects unknown capabilities
---
duration_ms: 2889.131
type: 'test'
...
# Subtest: request projection keeps exact tool pairs and complete recent user instructions while preserving disk history
ok 38 - request projection keeps exact tool pairs and complete recent user instructions while preserving disk history
---
duration_ms: 2106.8141
type: 'test'
...
# Subtest: a live SQLite host with the actual 1.9 artifact object upgrades, keeps its original blobs and verifies real browser PNG bytes
ok 39 - a live SQLite host with the actual 1.9 artifact object upgrades, keeps its original blobs and verifies real browser PNG bytes
---
duration_ms: 4369.1298
type: 'test'
...
# Subtest: failed upgrade restores the original artifact object, provider and request handler
ok 40 - failed upgrade restores the original artifact object, provider and request handler
---
duration_ms: 1701.0943
type: 'test'
...
# Subtest: claiming the same runtime owner twice preserves its original metadata and every database row
ok 41 - claiming the same runtime owner twice preserves its original metadata and every database row
---
duration_ms: 897.1056
type: 'test'
...
# Subtest: sessions without controlled-engine history retain their prior export shape and never expose another session record
ok 42 - sessions without controlled-engine history retain their prior export shape and never expose another session record
---
duration_ms: 1193.7581
type: 'test'
...
# Subtest: live workers, controlled drivers, or MCP operations refuse an upgrade before changing handlers, managers or database rows
ok 43 - live workers, controlled drivers, or MCP operations refuse an upgrade before changing handlers, managers or database rows
---
duration_ms: 969.9211
type: 'test'
...
# Subtest: an idle upgrade loads the newly introduced engine record map, resets obsolete managers and preserves original rows, sessions and provider objects
ok 44 - an idle upgrade loads the newly introduced engine record map, resets obsolete managers and preserves original rows, sessions and provider objects
---
duration_ms: 1010.0131
type: 'test'
...
# Subtest: an installation failure restores old managers, original record maps, handlers and all existing database rows
ok 45 - an installation failure restores old managers, original record maps, handlers and all existing database rows
---
duration_ms: 1579.1
type: 'test'
...
# Subtest: the owned controlled driver remains an active session and project even after outer task activity has ended
ok 46 - the owned controlled driver remains an active session and project even after outer task activity has ended
---
duration_ms: 1174.6617
type: 'test'
...
# Subtest: release and upgrade help or invalid arguments have no file or network setup side effects in an independent directory
ok 47 - release and upgrade help or invalid arguments have no file or network setup side effects in an independent directory
---
duration_ms: 1018.0031
type: 'test'
...
# Subtest: hot upgrade refreshes retained model and importer methods while preserving credential and routing objects
ok 48 - hot upgrade refreshes retained model and importer methods while preserving credential and routing objects
---
duration_ms: 674.2535
type: 'test'
...
# Subtest: hot upgrade refuses active probes and restores retained model methods if installation fails
ok 49 - hot upgrade refuses active probes and restores retained model methods if installation fails
---
duration_ms: 511.5755
type: 'test'
...
1..49
# tests 49
# suites 0
# pass 49
# fail 0
# cancelled 0
# skipped 0
# todo 0
# duration_ms 38427.6009
TAP version 13
# Subtest: host challenge proof rejects a stale challenge, changed identity and changed signature
ok 1 - host challenge proof rejects a stale challenge, changed identity and changed signature
---
duration_ms: 1.1348
type: 'test'
...
# Subtest: stop confirmation binds the original instance, creation and loaded runtime
ok 2 - stop confirmation binds the original instance, creation and loaded runtime
---
duration_ms: 0.1247
type: 'test'
...
# Subtest: state paths reject escaping, linked ancestors, ambiguous hardlinks and damaged JSON
ok 3 - state paths reject escaping, linked ancestors, ambiguous hardlinks and damaged JSON
---
duration_ms: 1133.2191
type: 'test'
...
# Subtest: failed registration publishes no initialized memory and can complete after its actual inputs are supplied
ok 4 - failed registration publishes no initialized memory and can complete after its actual inputs are supplied
---
duration_ms: 2526.6054
type: 'test'
...
# Subtest: host shutdown cancels an un-signalled HTTP model request and refuses later requests until recovered
ok 5 - host shutdown cancels an un-signalled HTTP model request and refuses later requests until recovered
---
duration_ms: 1.9739
type: 'test'
...
# Subtest: host shutdown cancels model-pool acquisition and releases queue ownership without inventing attempts
ok 6 - host shutdown cancels model-pool acquisition and releases queue ownership without inventing attempts
---
duration_ms: 818.2146
type: 'test'
...
# Subtest: a question about the previous workbench error receives local error context and current change states
ok 7 - a question about the previous workbench error receives local error context and current change states
---
duration_ms: 2213.6377
type: 'test'
...
# Subtest: a transient model failure after an approved command retries only the model request
ok 8 - a transient model failure after an approved command retries only the model request
---
duration_ms: 2266.197
type: 'test'
...
# Subtest: exhausted model retries preserve tool results, and explicit resume adds no new user turn or replay
ok 9 - exhausted model retries preserve tool results, and explicit resume adds no new user turn or replay
---
duration_ms: 3657.3963
type: 'test'
...
# Subtest: stopping during model backoff cancels further requests and clears the running request state
ok 10 - stopping during model backoff cancels further requests and clears the running request state
---
duration_ms: 1798.9338
type: 'test'
...
# Subtest: HTTP authentication errors are classified, redacted and never automatically retried
ok 11 - HTTP authentication errors are classified, redacted and never automatically retried
---
duration_ms: 21.9978
type: 'test'
...
# Subtest: temporary HTTP service errors retry and recover with recorded attempt counts
ok 12 - temporary HTTP service errors retry and recover with recorded attempt counts
---
duration_ms: 8.2405
type: 'test'
...
# Subtest: invalid and oversized protocol responses fail once with a useful response-format diagnosis
ok 13 - invalid and oversized protocol responses fail once with a useful response-format diagnosis
---
duration_ms: 4.0449
type: 'test'
...
# Subtest: HTTP rate limits expose their source and obey the service retry interval before recovery
ok 14 - HTTP rate limits expose their source and obey the service retry interval before recovery
---
duration_ms: 1010.8285
type: 'test'
...
# Subtest: Windows commands preserve Chinese through a PowerShell pipeline into Node
ok 15 - Windows commands preserve Chinese through a PowerShell pipeline into Node
---
duration_ms: 1003.6617
type: 'test'
...
# Subtest: Windows shell syntax errors stay readable and native command exit codes are retained
ok 16 - Windows shell syntax errors stay readable and native command exit codes are retained
---
duration_ms: 971.0757
type: 'test'
...
# Subtest: review and completed states do not turn a generic continue click into a new model request
ok 17 - review and completed states do not turn a generic continue click into a new model request
---
duration_ms: 852.3601
type: 'test'
...
# Subtest: old session data receives a truthful historical connection diagnosis without rewriting history
ok 18 - old session data receives a truthful historical connection diagnosis without rewriting history
---
duration_ms: 1178.3301
type: 'test'
...
# Subtest: model registry keeps original credentials/endpoint intact, rejects stale edits, and forgets only volatile keys on restart
ok 19 - model registry keeps original credentials/endpoint intact, rejects stale edits, and forgets only volatile keys on restart
---
duration_ms: 1356.0532
type: 'test'
...
# Subtest: config import previews exact revisions, keeps secrets volatile, syncs only when enabled and preserves manual edits on conflicts
ok 20 - config import previews exact revisions, keeps secrets volatile, syncs only when enabled and preserves manual edits on conflicts
---
duration_ms: 1132.7334
type: 'test'
...
# Subtest: real HTTP pool fails over 429 to another model and never repeats completed tools; attempts seal actual request bytes and Run identity
ok 21 - real HTTP pool fails over 429 to another model and never repeats completed tools; attempts seal actual request bytes and Run identity
---
duration_ms: 3159.0983
type: 'test'
...
# Subtest: pool rotation and weights select members on new Runs; disabled and missing credentials are excluded without requests
ok 22 - pool rotation and weights select members on new Runs; disabled and missing credentials are excluded without requests
---
duration_ms: 4730.0215
type: 'test'
...
# Subtest: 401 blocks the credential revision; updating the key restores availability without leaking error echoes
ok 23 - 401 blocks the credential revision; updating the key restores availability without leaking error echoes
---
duration_ms: 2035.7928
type: 'test'
...
# Subtest: queued request cancellation and in-flight disable release only owned leases and do not dispatch another HTTP request
ok 24 - queued request cancellation and in-flight disable release only owned leases and do not dispatch another HTTP request
---
duration_ms: 1053.5497
type: 'test'
...
# Subtest: compiled model-center HTTP API persists public profiles, blocks forged origins, and local operations never invoke a model
ok 25 - compiled model-center HTTP API persists public profiles, blocks forged origins, and local operations never invoke a model
---
duration_ms: 2488.7937
type: 'test'
...
# Subtest: model registry disk failures restore routing, pool profiles and credential references without partial updates
ok 26 - model registry disk failures restore routing, pool profiles and credential references without partial updates
---
duration_ms: 924.9926
type: 'test'
...
# Subtest: model selection and saved configuration commit together and failures preserve the original mode and endpoint
ok 27 - model selection and saved configuration commit together and failures preserve the original mode and endpoint
---
duration_ms: 896.6963
type: 'test'
...
# Subtest: configuration imports roll back the entire batch and retain one usable preview after validation or commit failure
ok 28 - configuration imports roll back the entire batch and retain one usable preview after validation or commit failure
---
duration_ms: 903.5305
type: 'test'
...
# Subtest: a stop from the event loop interrupts immediate model replies before later tools
ok 29 - a stop from the event loop interrupts immediate model replies before later tools
---
duration_ms: 1924.7977
type: 'test'
...
# Subtest: a stop after the first tool preserves a complete interrupted multi-call protocol group
ok 30 - a stop after the first tool preserves a complete interrupted multi-call protocol group
---
duration_ms: 2188.5101
type: 'test'
...
# Subtest: the compatible HTTP loop preserves call IDs, validates errors and waits before executing commands
ok 31 - the compatible HTTP loop preserves call IDs, validates errors and waits before executing commands
---
duration_ms: 5677.1928
type: 'test'
...
# Subtest: denial prevents command execution and produces a paired denied result
ok 32 - denial prevents command execution and produces a paired denied result
---
duration_ms: 1786.9452
type: 'test'
...
# Subtest: stopping during approval interrupts pending calls without executing or replaying them
ok 33 - stopping during approval interrupts pending calls without executing or replaying them
---
duration_ms: 1887.1945
type: 'test'
...
# Subtest: stopping a running fixture command terminates its own process tree and retains uncertain state
ok 34 - stopping a running fixture command terminates its own process tree and retains uncertain state
---
duration_ms: 1887.357
type: 'test'
...
# Subtest: command output decodes split UTF-8 and excludes the workbench key from child environment
ok 35 - command output decodes split UTF-8 and excludes the workbench key from child environment
---
duration_ms: 1547.6495
type: 'test'
...
# Subtest: model-list testing uses a tentative key without changing the saved connection and redacts service errors
ok 36 - model-list testing uses a tentative key without changing the saved connection and redacts service errors
---
duration_ms: 6.3458
type: 'test'
...
# Subtest: truncated or repeated model calls fail before execution; tool input validation rejects unknown capabilities
ok 37 - truncated or repeated model calls fail before execution; tool input validation rejects unknown capabilities
---
duration_ms: 2889.131
type: 'test'
...
# Subtest: request projection keeps exact tool pairs and complete recent user instructions while preserving disk history
ok 38 - request projection keeps exact tool pairs and complete recent user instructions while preserving disk history
---
duration_ms: 2106.8141
type: 'test'
...
# Subtest: a live SQLite host with the actual 1.9 artifact object upgrades, keeps its original blobs and verifies real browser PNG bytes
ok 39 - a live SQLite host with the actual 1.9 artifact object upgrades, keeps its original blobs and verifies real browser PNG bytes
---
duration_ms: 4369.1298
type: 'test'
...
# Subtest: failed upgrade restores the original artifact object, provider and request handler
ok 40 - failed upgrade restores the original artifact object, provider and request handler
---
duration_ms: 1701.0943
type: 'test'
...
# Subtest: claiming the same runtime owner twice preserves its original metadata and every database row
ok 41 - claiming the same runtime owner twice preserves its original metadata and every database row
---
duration_ms: 897.1056
type: 'test'
...
# Subtest: sessions without controlled-engine history retain their prior export shape and never expose another session record
ok 42 - sessions without controlled-engine history retain their prior export shape and never expose another session record
---
duration_ms: 1193.7581
type: 'test'
...
# Subtest: live workers, controlled drivers, or MCP operations refuse an upgrade before changing handlers, managers or database rows
ok 43 - live workers, controlled drivers, or MCP operations refuse an upgrade before changing handlers, managers or database rows
---
duration_ms: 969.9211
type: 'test'
...
# Subtest: an idle upgrade loads the newly introduced engine record map, resets obsolete managers and preserves original rows, sessions and provider objects
ok 44 - an idle upgrade loads the newly introduced engine record map, resets obsolete managers and preserves original rows, sessions and provider objects
---
duration_ms: 1010.0131
type: 'test'
...
# Subtest: an installation failure restores old managers, original record maps, handlers and all existing database rows
ok 45 - an installation failure restores old managers, original record maps, handlers and all existing database rows
---
duration_ms: 1579.1
type: 'test'
...
# Subtest: the owned controlled driver remains an active session and project even after outer task activity has ended
ok 46 - the owned controlled driver remains an active session and project even after outer task activity has ended
---
duration_ms: 1174.6617
type: 'test'
...
# Subtest: release and upgrade help or invalid arguments have no file or network setup side effects in an independent directory
ok 47 - release and upgrade help or invalid arguments have no file or network setup side effects in an independent directory
---
duration_ms: 1018.0031
type: 'test'
...
# Subtest: hot upgrade refreshes retained model and importer methods while preserving credential and routing objects
ok 48 - hot upgrade refreshes retained model and importer methods while preserving credential and routing objects
---
duration_ms: 674.2535
type: 'test'
...
# Subtest: hot upgrade refuses active probes and restores retained model methods if installation fails
ok 49 - hot upgrade refuses active probes and restores retained model methods if installation fails
---
duration_ms: 511.5755
type: 'test'
...
1..49
# tests 49
# suites 0
# pass 49
# fail 0
# cancelled 0
# skipped 0
# todo 0
# duration_ms 38427.6009
+319 -319
View File
@@ -1,319 +1,319 @@
TAP version 13
# Subtest: host challenge proof rejects a stale challenge, changed identity and changed signature
ok 1 - host challenge proof rejects a stale challenge, changed identity and changed signature
---
duration_ms: 1.2329
type: 'test'
...
# Subtest: stop confirmation binds the original instance, creation and loaded runtime
ok 2 - stop confirmation binds the original instance, creation and loaded runtime
---
duration_ms: 0.1251
type: 'test'
...
# Subtest: state paths reject escaping, linked ancestors, ambiguous hardlinks and damaged JSON
ok 3 - state paths reject escaping, linked ancestors, ambiguous hardlinks and damaged JSON
---
duration_ms: 1235.5945
type: 'test'
...
# Subtest: failed registration publishes no initialized memory and can complete after its actual inputs are supplied
not ok 4 - failed registration publishes no initialized memory and can complete after its actual inputs are supplied
---
duration_ms: 2624.3439
type: 'test'
location: 'D:\\Ai-project\\REA逆向\\Agent-Workbench\\tests\\desktop-host.test.mjs:40:1'
failureType: 'testCodeFailure'
error: '无法为本机秘密目录设置当前Windows用户权限。'
code: 'HOST_DIRECTORY_UNPROTECTED'
stack: |-
hostError (file:///D:/Ai-project/REA%E9%80%86%E5%90%91/Agent-Workbench/src/shared/contracts/desktop-host.mjs:5:100)
protectUserDirectory (file:///D:/Ai-project/REA%E9%80%86%E5%90%91/Agent-Workbench/src/core/desktop/process-identity.mjs:21:17)
HostAuthority.initialize (file:///D:/Ai-project/REA%E9%80%86%E5%90%91/Agent-Workbench/src/core/desktop/host-authority.mjs:18:5)
TestContext.<anonymous> (file:///D:/Ai-project/REA%E9%80%86%E5%90%91/Agent-Workbench/tests/desktop-host.test.mjs:46:13)
Test.runInAsyncScope (node:async_hooks:227:14)
Test.run (node:internal/test_runner/test:1402:25)
Test.processPendingSubtests (node:internal/test_runner/test:974:18)
Test.postRun (node:internal/test_runner/test:1542:19)
Test.run (node:internal/test_runner/test:1467:12)
async Test.processPendingSubtests (node:internal/test_runner/test:974:7)
...
# Subtest: host shutdown cancels an un-signalled HTTP model request and refuses later requests until recovered
ok 5 - host shutdown cancels an un-signalled HTTP model request and refuses later requests until recovered
---
duration_ms: 2.5545
type: 'test'
...
# Subtest: host shutdown cancels model-pool acquisition and releases queue ownership without inventing attempts
ok 6 - host shutdown cancels model-pool acquisition and releases queue ownership without inventing attempts
---
duration_ms: 1009.2534
type: 'test'
...
# Subtest: a question about the previous workbench error receives local error context and current change states
ok 7 - a question about the previous workbench error receives local error context and current change states
---
duration_ms: 2355.5938
type: 'test'
...
# Subtest: a transient model failure after an approved command retries only the model request
ok 8 - a transient model failure after an approved command retries only the model request
---
duration_ms: 2540.0191
type: 'test'
...
# Subtest: exhausted model retries preserve tool results, and explicit resume adds no new user turn or replay
ok 9 - exhausted model retries preserve tool results, and explicit resume adds no new user turn or replay
---
duration_ms: 3629.734
type: 'test'
...
# Subtest: stopping during model backoff cancels further requests and clears the running request state
ok 10 - stopping during model backoff cancels further requests and clears the running request state
---
duration_ms: 1609.2816
type: 'test'
...
# Subtest: HTTP authentication errors are classified, redacted and never automatically retried
ok 11 - HTTP authentication errors are classified, redacted and never automatically retried
---
duration_ms: 21.1158
type: 'test'
...
# Subtest: temporary HTTP service errors retry and recover with recorded attempt counts
ok 12 - temporary HTTP service errors retry and recover with recorded attempt counts
---
duration_ms: 6.9441
type: 'test'
...
# Subtest: invalid and oversized protocol responses fail once with a useful response-format diagnosis
ok 13 - invalid and oversized protocol responses fail once with a useful response-format diagnosis
---
duration_ms: 4.0345
type: 'test'
...
# Subtest: HTTP rate limits expose their source and obey the service retry interval before recovery
ok 14 - HTTP rate limits expose their source and obey the service retry interval before recovery
---
duration_ms: 1022.397
type: 'test'
...
# Subtest: Windows commands preserve Chinese through a PowerShell pipeline into Node
ok 15 - Windows commands preserve Chinese through a PowerShell pipeline into Node
---
duration_ms: 1069.2031
type: 'test'
...
# Subtest: Windows shell syntax errors stay readable and native command exit codes are retained
ok 16 - Windows shell syntax errors stay readable and native command exit codes are retained
---
duration_ms: 1045.5572
type: 'test'
...
# Subtest: review and completed states do not turn a generic continue click into a new model request
ok 17 - review and completed states do not turn a generic continue click into a new model request
---
duration_ms: 920.9875
type: 'test'
...
# Subtest: old session data receives a truthful historical connection diagnosis without rewriting history
ok 18 - old session data receives a truthful historical connection diagnosis without rewriting history
---
duration_ms: 919.6361
type: 'test'
...
# Subtest: model registry keeps original credentials/endpoint intact, rejects stale edits, and forgets only volatile keys on restart
ok 19 - model registry keeps original credentials/endpoint intact, rejects stale edits, and forgets only volatile keys on restart
---
duration_ms: 1320.4966
type: 'test'
...
# Subtest: config import previews exact revisions, keeps secrets volatile, syncs only when enabled and preserves manual edits on conflicts
ok 20 - config import previews exact revisions, keeps secrets volatile, syncs only when enabled and preserves manual edits on conflicts
---
duration_ms: 1097.0453
type: 'test'
...
# Subtest: real HTTP pool fails over 429 to another model and never repeats completed tools; attempts seal actual request bytes and Run identity
ok 21 - real HTTP pool fails over 429 to another model and never repeats completed tools; attempts seal actual request bytes and Run identity
---
duration_ms: 3015.9964
type: 'test'
...
# Subtest: pool rotation and weights select members on new Runs; disabled and missing credentials are excluded without requests
ok 22 - pool rotation and weights select members on new Runs; disabled and missing credentials are excluded without requests
---
duration_ms: 4841.3906
type: 'test'
...
# Subtest: 401 blocks the credential revision; updating the key restores availability without leaking error echoes
ok 23 - 401 blocks the credential revision; updating the key restores availability without leaking error echoes
---
duration_ms: 2150.3254
type: 'test'
...
# Subtest: queued request cancellation and in-flight disable release only owned leases and do not dispatch another HTTP request
ok 24 - queued request cancellation and in-flight disable release only owned leases and do not dispatch another HTTP request
---
duration_ms: 1305.1053
type: 'test'
...
# Subtest: compiled model-center HTTP API persists public profiles, blocks forged origins, and local operations never invoke a model
ok 25 - compiled model-center HTTP API persists public profiles, blocks forged origins, and local operations never invoke a model
---
duration_ms: 2560.933
type: 'test'
...
# Subtest: model registry disk failures restore routing, pool profiles and credential references without partial updates
ok 26 - model registry disk failures restore routing, pool profiles and credential references without partial updates
---
duration_ms: 959.2805
type: 'test'
...
# Subtest: model selection and saved configuration commit together and failures preserve the original mode and endpoint
ok 27 - model selection and saved configuration commit together and failures preserve the original mode and endpoint
---
duration_ms: 976.065
type: 'test'
...
# Subtest: configuration imports roll back the entire batch and retain one usable preview after validation or commit failure
ok 28 - configuration imports roll back the entire batch and retain one usable preview after validation or commit failure
---
duration_ms: 833.4502
type: 'test'
...
# Subtest: a stop from the event loop interrupts immediate model replies before later tools
ok 29 - a stop from the event loop interrupts immediate model replies before later tools
---
duration_ms: 2200.8626
type: 'test'
...
# Subtest: a stop after the first tool preserves a complete interrupted multi-call protocol group
ok 30 - a stop after the first tool preserves a complete interrupted multi-call protocol group
---
duration_ms: 2434.7655
type: 'test'
...
# Subtest: the compatible HTTP loop preserves call IDs, validates errors and waits before executing commands
ok 31 - the compatible HTTP loop preserves call IDs, validates errors and waits before executing commands
---
duration_ms: 5449.6749
type: 'test'
...
# Subtest: denial prevents command execution and produces a paired denied result
ok 32 - denial prevents command execution and produces a paired denied result
---
duration_ms: 1648.3382
type: 'test'
...
# Subtest: stopping during approval interrupts pending calls without executing or replaying them
ok 33 - stopping during approval interrupts pending calls without executing or replaying them
---
duration_ms: 1643.657
type: 'test'
...
# Subtest: stopping a running fixture command terminates its own process tree and retains uncertain state
ok 34 - stopping a running fixture command terminates its own process tree and retains uncertain state
---
duration_ms: 2162.5898
type: 'test'
...
# Subtest: command output decodes split UTF-8 and excludes the workbench key from child environment
ok 35 - command output decodes split UTF-8 and excludes the workbench key from child environment
---
duration_ms: 1329.7417
type: 'test'
...
# Subtest: model-list testing uses a tentative key without changing the saved connection and redacts service errors
ok 36 - model-list testing uses a tentative key without changing the saved connection and redacts service errors
---
duration_ms: 6.2866
type: 'test'
...
# Subtest: truncated or repeated model calls fail before execution; tool input validation rejects unknown capabilities
ok 37 - truncated or repeated model calls fail before execution; tool input validation rejects unknown capabilities
---
duration_ms: 3618.0824
type: 'test'
...
# Subtest: request projection keeps exact tool pairs and complete recent user instructions while preserving disk history
ok 38 - request projection keeps exact tool pairs and complete recent user instructions while preserving disk history
---
duration_ms: 1269.4481
type: 'test'
...
# Subtest: a live SQLite host with the actual 1.9 artifact object upgrades, keeps its original blobs and verifies real browser PNG bytes
ok 39 - a live SQLite host with the actual 1.9 artifact object upgrades, keeps its original blobs and verifies real browser PNG bytes
---
duration_ms: 4653.6582
type: 'test'
...
# Subtest: failed upgrade restores the original artifact object, provider and request handler
ok 40 - failed upgrade restores the original artifact object, provider and request handler
---
duration_ms: 2039.9906
type: 'test'
...
# Subtest: claiming the same runtime owner twice preserves its original metadata and every database row
ok 41 - claiming the same runtime owner twice preserves its original metadata and every database row
---
duration_ms: 1004.9648
type: 'test'
...
# Subtest: sessions without controlled-engine history retain their prior export shape and never expose another session record
ok 42 - sessions without controlled-engine history retain their prior export shape and never expose another session record
---
duration_ms: 994.6092
type: 'test'
...
# Subtest: live workers, controlled drivers, or MCP operations refuse an upgrade before changing handlers, managers or database rows
ok 43 - live workers, controlled drivers, or MCP operations refuse an upgrade before changing handlers, managers or database rows
---
duration_ms: 897.7159
type: 'test'
...
# Subtest: an idle upgrade loads the newly introduced engine record map, resets obsolete managers and preserves original rows, sessions and provider objects
ok 44 - an idle upgrade loads the newly introduced engine record map, resets obsolete managers and preserves original rows, sessions and provider objects
---
duration_ms: 1065.9033
type: 'test'
...
# Subtest: an installation failure restores old managers, original record maps, handlers and all existing database rows
ok 45 - an installation failure restores old managers, original record maps, handlers and all existing database rows
---
duration_ms: 1815.9839
type: 'test'
...
# Subtest: the owned controlled driver remains an active session and project even after outer task activity has ended
ok 46 - the owned controlled driver remains an active session and project even after outer task activity has ended
---
duration_ms: 1021.041
type: 'test'
...
# Subtest: release and upgrade help or invalid arguments have no file or network setup side effects in an independent directory
ok 47 - release and upgrade help or invalid arguments have no file or network setup side effects in an independent directory
---
duration_ms: 1169.9017
type: 'test'
...
# Subtest: hot upgrade refreshes retained model and importer methods while preserving credential and routing objects
ok 48 - hot upgrade refreshes retained model and importer methods while preserving credential and routing objects
---
duration_ms: 982.3635
type: 'test'
...
# Subtest: hot upgrade refuses active probes and restores retained model methods if installation fails
ok 49 - hot upgrade refuses active probes and restores retained model methods if installation fails
---
duration_ms: 938.8454
type: 'test'
...
1..49
# tests 49
# suites 0
# pass 48
# fail 1
# cancelled 0
# skipped 0
# todo 0
# duration_ms 39180.5899
TAP version 13
# Subtest: host challenge proof rejects a stale challenge, changed identity and changed signature
ok 1 - host challenge proof rejects a stale challenge, changed identity and changed signature
---
duration_ms: 1.2329
type: 'test'
...
# Subtest: stop confirmation binds the original instance, creation and loaded runtime
ok 2 - stop confirmation binds the original instance, creation and loaded runtime
---
duration_ms: 0.1251
type: 'test'
...
# Subtest: state paths reject escaping, linked ancestors, ambiguous hardlinks and damaged JSON
ok 3 - state paths reject escaping, linked ancestors, ambiguous hardlinks and damaged JSON
---
duration_ms: 1235.5945
type: 'test'
...
# Subtest: failed registration publishes no initialized memory and can complete after its actual inputs are supplied
not ok 4 - failed registration publishes no initialized memory and can complete after its actual inputs are supplied
---
duration_ms: 2624.3439
type: 'test'
location: 'D:\\Ai-project\\REA逆向\\Agent-Workbench\\tests\\desktop-host.test.mjs:40:1'
failureType: 'testCodeFailure'
error: '无法为本机秘密目录设置当前Windows用户权限。'
code: 'HOST_DIRECTORY_UNPROTECTED'
stack: |-
hostError (file:///D:/Ai-project/REA%E9%80%86%E5%90%91/Agent-Workbench/src/shared/contracts/desktop-host.mjs:5:100)
protectUserDirectory (file:///D:/Ai-project/REA%E9%80%86%E5%90%91/Agent-Workbench/src/core/desktop/process-identity.mjs:21:17)
HostAuthority.initialize (file:///D:/Ai-project/REA%E9%80%86%E5%90%91/Agent-Workbench/src/core/desktop/host-authority.mjs:18:5)
TestContext.<anonymous> (file:///D:/Ai-project/REA%E9%80%86%E5%90%91/Agent-Workbench/tests/desktop-host.test.mjs:46:13)
Test.runInAsyncScope (node:async_hooks:227:14)
Test.run (node:internal/test_runner/test:1402:25)
Test.processPendingSubtests (node:internal/test_runner/test:974:18)
Test.postRun (node:internal/test_runner/test:1542:19)
Test.run (node:internal/test_runner/test:1467:12)
async Test.processPendingSubtests (node:internal/test_runner/test:974:7)
...
# Subtest: host shutdown cancels an un-signalled HTTP model request and refuses later requests until recovered
ok 5 - host shutdown cancels an un-signalled HTTP model request and refuses later requests until recovered
---
duration_ms: 2.5545
type: 'test'
...
# Subtest: host shutdown cancels model-pool acquisition and releases queue ownership without inventing attempts
ok 6 - host shutdown cancels model-pool acquisition and releases queue ownership without inventing attempts
---
duration_ms: 1009.2534
type: 'test'
...
# Subtest: a question about the previous workbench error receives local error context and current change states
ok 7 - a question about the previous workbench error receives local error context and current change states
---
duration_ms: 2355.5938
type: 'test'
...
# Subtest: a transient model failure after an approved command retries only the model request
ok 8 - a transient model failure after an approved command retries only the model request
---
duration_ms: 2540.0191
type: 'test'
...
# Subtest: exhausted model retries preserve tool results, and explicit resume adds no new user turn or replay
ok 9 - exhausted model retries preserve tool results, and explicit resume adds no new user turn or replay
---
duration_ms: 3629.734
type: 'test'
...
# Subtest: stopping during model backoff cancels further requests and clears the running request state
ok 10 - stopping during model backoff cancels further requests and clears the running request state
---
duration_ms: 1609.2816
type: 'test'
...
# Subtest: HTTP authentication errors are classified, redacted and never automatically retried
ok 11 - HTTP authentication errors are classified, redacted and never automatically retried
---
duration_ms: 21.1158
type: 'test'
...
# Subtest: temporary HTTP service errors retry and recover with recorded attempt counts
ok 12 - temporary HTTP service errors retry and recover with recorded attempt counts
---
duration_ms: 6.9441
type: 'test'
...
# Subtest: invalid and oversized protocol responses fail once with a useful response-format diagnosis
ok 13 - invalid and oversized protocol responses fail once with a useful response-format diagnosis
---
duration_ms: 4.0345
type: 'test'
...
# Subtest: HTTP rate limits expose their source and obey the service retry interval before recovery
ok 14 - HTTP rate limits expose their source and obey the service retry interval before recovery
---
duration_ms: 1022.397
type: 'test'
...
# Subtest: Windows commands preserve Chinese through a PowerShell pipeline into Node
ok 15 - Windows commands preserve Chinese through a PowerShell pipeline into Node
---
duration_ms: 1069.2031
type: 'test'
...
# Subtest: Windows shell syntax errors stay readable and native command exit codes are retained
ok 16 - Windows shell syntax errors stay readable and native command exit codes are retained
---
duration_ms: 1045.5572
type: 'test'
...
# Subtest: review and completed states do not turn a generic continue click into a new model request
ok 17 - review and completed states do not turn a generic continue click into a new model request
---
duration_ms: 920.9875
type: 'test'
...
# Subtest: old session data receives a truthful historical connection diagnosis without rewriting history
ok 18 - old session data receives a truthful historical connection diagnosis without rewriting history
---
duration_ms: 919.6361
type: 'test'
...
# Subtest: model registry keeps original credentials/endpoint intact, rejects stale edits, and forgets only volatile keys on restart
ok 19 - model registry keeps original credentials/endpoint intact, rejects stale edits, and forgets only volatile keys on restart
---
duration_ms: 1320.4966
type: 'test'
...
# Subtest: config import previews exact revisions, keeps secrets volatile, syncs only when enabled and preserves manual edits on conflicts
ok 20 - config import previews exact revisions, keeps secrets volatile, syncs only when enabled and preserves manual edits on conflicts
---
duration_ms: 1097.0453
type: 'test'
...
# Subtest: real HTTP pool fails over 429 to another model and never repeats completed tools; attempts seal actual request bytes and Run identity
ok 21 - real HTTP pool fails over 429 to another model and never repeats completed tools; attempts seal actual request bytes and Run identity
---
duration_ms: 3015.9964
type: 'test'
...
# Subtest: pool rotation and weights select members on new Runs; disabled and missing credentials are excluded without requests
ok 22 - pool rotation and weights select members on new Runs; disabled and missing credentials are excluded without requests
---
duration_ms: 4841.3906
type: 'test'
...
# Subtest: 401 blocks the credential revision; updating the key restores availability without leaking error echoes
ok 23 - 401 blocks the credential revision; updating the key restores availability without leaking error echoes
---
duration_ms: 2150.3254
type: 'test'
...
# Subtest: queued request cancellation and in-flight disable release only owned leases and do not dispatch another HTTP request
ok 24 - queued request cancellation and in-flight disable release only owned leases and do not dispatch another HTTP request
---
duration_ms: 1305.1053
type: 'test'
...
# Subtest: compiled model-center HTTP API persists public profiles, blocks forged origins, and local operations never invoke a model
ok 25 - compiled model-center HTTP API persists public profiles, blocks forged origins, and local operations never invoke a model
---
duration_ms: 2560.933
type: 'test'
...
# Subtest: model registry disk failures restore routing, pool profiles and credential references without partial updates
ok 26 - model registry disk failures restore routing, pool profiles and credential references without partial updates
---
duration_ms: 959.2805
type: 'test'
...
# Subtest: model selection and saved configuration commit together and failures preserve the original mode and endpoint
ok 27 - model selection and saved configuration commit together and failures preserve the original mode and endpoint
---
duration_ms: 976.065
type: 'test'
...
# Subtest: configuration imports roll back the entire batch and retain one usable preview after validation or commit failure
ok 28 - configuration imports roll back the entire batch and retain one usable preview after validation or commit failure
---
duration_ms: 833.4502
type: 'test'
...
# Subtest: a stop from the event loop interrupts immediate model replies before later tools
ok 29 - a stop from the event loop interrupts immediate model replies before later tools
---
duration_ms: 2200.8626
type: 'test'
...
# Subtest: a stop after the first tool preserves a complete interrupted multi-call protocol group
ok 30 - a stop after the first tool preserves a complete interrupted multi-call protocol group
---
duration_ms: 2434.7655
type: 'test'
...
# Subtest: the compatible HTTP loop preserves call IDs, validates errors and waits before executing commands
ok 31 - the compatible HTTP loop preserves call IDs, validates errors and waits before executing commands
---
duration_ms: 5449.6749
type: 'test'
...
# Subtest: denial prevents command execution and produces a paired denied result
ok 32 - denial prevents command execution and produces a paired denied result
---
duration_ms: 1648.3382
type: 'test'
...
# Subtest: stopping during approval interrupts pending calls without executing or replaying them
ok 33 - stopping during approval interrupts pending calls without executing or replaying them
---
duration_ms: 1643.657
type: 'test'
...
# Subtest: stopping a running fixture command terminates its own process tree and retains uncertain state
ok 34 - stopping a running fixture command terminates its own process tree and retains uncertain state
---
duration_ms: 2162.5898
type: 'test'
...
# Subtest: command output decodes split UTF-8 and excludes the workbench key from child environment
ok 35 - command output decodes split UTF-8 and excludes the workbench key from child environment
---
duration_ms: 1329.7417
type: 'test'
...
# Subtest: model-list testing uses a tentative key without changing the saved connection and redacts service errors
ok 36 - model-list testing uses a tentative key without changing the saved connection and redacts service errors
---
duration_ms: 6.2866
type: 'test'
...
# Subtest: truncated or repeated model calls fail before execution; tool input validation rejects unknown capabilities
ok 37 - truncated or repeated model calls fail before execution; tool input validation rejects unknown capabilities
---
duration_ms: 3618.0824
type: 'test'
...
# Subtest: request projection keeps exact tool pairs and complete recent user instructions while preserving disk history
ok 38 - request projection keeps exact tool pairs and complete recent user instructions while preserving disk history
---
duration_ms: 1269.4481
type: 'test'
...
# Subtest: a live SQLite host with the actual 1.9 artifact object upgrades, keeps its original blobs and verifies real browser PNG bytes
ok 39 - a live SQLite host with the actual 1.9 artifact object upgrades, keeps its original blobs and verifies real browser PNG bytes
---
duration_ms: 4653.6582
type: 'test'
...
# Subtest: failed upgrade restores the original artifact object, provider and request handler
ok 40 - failed upgrade restores the original artifact object, provider and request handler
---
duration_ms: 2039.9906
type: 'test'
...
# Subtest: claiming the same runtime owner twice preserves its original metadata and every database row
ok 41 - claiming the same runtime owner twice preserves its original metadata and every database row
---
duration_ms: 1004.9648
type: 'test'
...
# Subtest: sessions without controlled-engine history retain their prior export shape and never expose another session record
ok 42 - sessions without controlled-engine history retain their prior export shape and never expose another session record
---
duration_ms: 994.6092
type: 'test'
...
# Subtest: live workers, controlled drivers, or MCP operations refuse an upgrade before changing handlers, managers or database rows
ok 43 - live workers, controlled drivers, or MCP operations refuse an upgrade before changing handlers, managers or database rows
---
duration_ms: 897.7159
type: 'test'
...
# Subtest: an idle upgrade loads the newly introduced engine record map, resets obsolete managers and preserves original rows, sessions and provider objects
ok 44 - an idle upgrade loads the newly introduced engine record map, resets obsolete managers and preserves original rows, sessions and provider objects
---
duration_ms: 1065.9033
type: 'test'
...
# Subtest: an installation failure restores old managers, original record maps, handlers and all existing database rows
ok 45 - an installation failure restores old managers, original record maps, handlers and all existing database rows
---
duration_ms: 1815.9839
type: 'test'
...
# Subtest: the owned controlled driver remains an active session and project even after outer task activity has ended
ok 46 - the owned controlled driver remains an active session and project even after outer task activity has ended
---
duration_ms: 1021.041
type: 'test'
...
# Subtest: release and upgrade help or invalid arguments have no file or network setup side effects in an independent directory
ok 47 - release and upgrade help or invalid arguments have no file or network setup side effects in an independent directory
---
duration_ms: 1169.9017
type: 'test'
...
# Subtest: hot upgrade refreshes retained model and importer methods while preserving credential and routing objects
ok 48 - hot upgrade refreshes retained model and importer methods while preserving credential and routing objects
---
duration_ms: 982.3635
type: 'test'
...
# Subtest: hot upgrade refuses active probes and restores retained model methods if installation fails
ok 49 - hot upgrade refuses active probes and restores retained model methods if installation fails
---
duration_ms: 938.8454
type: 'test'
...
1..49
# tests 49
# suites 0
# pass 48
# fail 1
# cancelled 0
# skipped 0
# todo 0
# duration_ms 39180.5899
+9 -9
View File
@@ -1,9 +1,9 @@
concurrent start: passed
independent creation and user observation: passed
wrong data root and installation rejected: passed
stale registration and creation rejected: passed
damaged service index rejected: passed
running input binding does not silently change: passed
foreign health service rejected: passed
completed tools and approvals retained on stop and restart: passed
actual model HTTP request abort and verified process exit: passed
concurrent start: passed
independent creation and user observation: passed
wrong data root and installation rejected: passed
stale registration and creation rejected: passed
damaged service index rejected: passed
running input binding does not silently change: passed
foreign health service rejected: passed
completed tools and approvals retained on stop and restart: passed
actual model HTTP request abort and verified process exit: passed
+148 -148
View File
@@ -1,148 +1,148 @@
TAP version 13
# Subtest: host challenge proof rejects a stale challenge, changed identity and changed signature
ok 1 - host challenge proof rejects a stale challenge, changed identity and changed signature
---
duration_ms: 1.1912
type: 'test'
...
# Subtest: stop confirmation binds the original instance, creation and loaded runtime
ok 2 - stop confirmation binds the original instance, creation and loaded runtime
---
duration_ms: 0.1222
type: 'test'
...
# Subtest: state paths reject escaping, linked ancestors, ambiguous hardlinks and damaged JSON
ok 3 - state paths reject escaping, linked ancestors, ambiguous hardlinks and damaged JSON
---
duration_ms: 1246.7954
type: 'test'
...
# Subtest: failed registration publishes no initialized memory and can complete after its actual inputs are supplied
ok 4 - failed registration publishes no initialized memory and can complete after its actual inputs are supplied
---
duration_ms: 2518.2715
type: 'test'
...
# Subtest: host shutdown cancels an un-signalled HTTP model request and refuses later requests until recovered
ok 5 - host shutdown cancels an un-signalled HTTP model request and refuses later requests until recovered
---
duration_ms: 1.348
type: 'test'
...
# Subtest: host shutdown cancels model-pool acquisition and releases queue ownership without inventing attempts
ok 6 - host shutdown cancels model-pool acquisition and releases queue ownership without inventing attempts
---
duration_ms: 1092.8051
type: 'test'
...
# Subtest: model registry keeps original credentials/endpoint intact, rejects stale edits, and forgets only volatile keys on restart
ok 7 - model registry keeps original credentials/endpoint intact, rejects stale edits, and forgets only volatile keys on restart
---
duration_ms: 1416.2692
type: 'test'
...
# Subtest: config import previews exact revisions, keeps secrets volatile, syncs only when enabled and preserves manual edits on conflicts
ok 8 - config import previews exact revisions, keeps secrets volatile, syncs only when enabled and preserves manual edits on conflicts
---
duration_ms: 990.0314
type: 'test'
...
# Subtest: real HTTP pool fails over 429 to another model and never repeats completed tools; attempts seal actual request bytes and Run identity
ok 9 - real HTTP pool fails over 429 to another model and never repeats completed tools; attempts seal actual request bytes and Run identity
---
duration_ms: 3079.9908
type: 'test'
...
# Subtest: pool rotation and weights select members on new Runs; disabled and missing credentials are excluded without requests
ok 10 - pool rotation and weights select members on new Runs; disabled and missing credentials are excluded without requests
---
duration_ms: 5018.7778
type: 'test'
...
# Subtest: 401 blocks the credential revision; updating the key restores availability without leaking error echoes
ok 11 - 401 blocks the credential revision; updating the key restores availability without leaking error echoes
---
duration_ms: 1973.4865
type: 'test'
...
# Subtest: queued request cancellation and in-flight disable release only owned leases and do not dispatch another HTTP request
ok 12 - queued request cancellation and in-flight disable release only owned leases and do not dispatch another HTTP request
---
duration_ms: 963.6111
type: 'test'
...
# Subtest: compiled model-center HTTP API persists public profiles, blocks forged origins, and local operations never invoke a model
ok 13 - compiled model-center HTTP API persists public profiles, blocks forged origins, and local operations never invoke a model
---
duration_ms: 2006.8513
type: 'test'
...
# Subtest: model registry disk failures restore routing, pool profiles and credential references without partial updates
ok 14 - model registry disk failures restore routing, pool profiles and credential references without partial updates
---
duration_ms: 566.9346
type: 'test'
...
# Subtest: model selection and saved configuration commit together and failures preserve the original mode and endpoint
ok 15 - model selection and saved configuration commit together and failures preserve the original mode and endpoint
---
duration_ms: 693.7432
type: 'test'
...
# Subtest: configuration imports roll back the entire batch and retain one usable preview after validation or commit failure
ok 16 - configuration imports roll back the entire batch and retain one usable preview after validation or commit failure
---
duration_ms: 574.4074
type: 'test'
...
# Subtest: claiming the same runtime owner twice preserves its original metadata and every database row
ok 17 - claiming the same runtime owner twice preserves its original metadata and every database row
---
duration_ms: 1095.9592
type: 'test'
...
# Subtest: sessions without controlled-engine history retain their prior export shape and never expose another session record
ok 18 - sessions without controlled-engine history retain their prior export shape and never expose another session record
---
duration_ms: 995.0958
type: 'test'
...
# Subtest: live workers, controlled drivers, or MCP operations refuse an upgrade before changing handlers, managers or database rows
ok 19 - live workers, controlled drivers, or MCP operations refuse an upgrade before changing handlers, managers or database rows
---
duration_ms: 949.7472
type: 'test'
...
# Subtest: an idle upgrade loads the newly introduced engine record map, resets obsolete managers and preserves original rows, sessions and provider objects
ok 20 - an idle upgrade loads the newly introduced engine record map, resets obsolete managers and preserves original rows, sessions and provider objects
---
duration_ms: 1051.4217
type: 'test'
...
# Subtest: an installation failure restores old managers, original record maps, handlers and all existing database rows
ok 21 - an installation failure restores old managers, original record maps, handlers and all existing database rows
---
duration_ms: 1565.5388
type: 'test'
...
# Subtest: the owned controlled driver remains an active session and project even after outer task activity has ended
ok 22 - the owned controlled driver remains an active session and project even after outer task activity has ended
---
duration_ms: 1000.6998
type: 'test'
...
# Subtest: release and upgrade help or invalid arguments have no file or network setup side effects in an independent directory
ok 23 - release and upgrade help or invalid arguments have no file or network setup side effects in an independent directory
---
duration_ms: 990.8517
type: 'test'
...
1..23
# tests 23
# suites 0
# pass 23
# fail 0
# cancelled 0
# skipped 0
# todo 0
# duration_ms 17409.5834
TAP version 13
# Subtest: host challenge proof rejects a stale challenge, changed identity and changed signature
ok 1 - host challenge proof rejects a stale challenge, changed identity and changed signature
---
duration_ms: 1.1912
type: 'test'
...
# Subtest: stop confirmation binds the original instance, creation and loaded runtime
ok 2 - stop confirmation binds the original instance, creation and loaded runtime
---
duration_ms: 0.1222
type: 'test'
...
# Subtest: state paths reject escaping, linked ancestors, ambiguous hardlinks and damaged JSON
ok 3 - state paths reject escaping, linked ancestors, ambiguous hardlinks and damaged JSON
---
duration_ms: 1246.7954
type: 'test'
...
# Subtest: failed registration publishes no initialized memory and can complete after its actual inputs are supplied
ok 4 - failed registration publishes no initialized memory and can complete after its actual inputs are supplied
---
duration_ms: 2518.2715
type: 'test'
...
# Subtest: host shutdown cancels an un-signalled HTTP model request and refuses later requests until recovered
ok 5 - host shutdown cancels an un-signalled HTTP model request and refuses later requests until recovered
---
duration_ms: 1.348
type: 'test'
...
# Subtest: host shutdown cancels model-pool acquisition and releases queue ownership without inventing attempts
ok 6 - host shutdown cancels model-pool acquisition and releases queue ownership without inventing attempts
---
duration_ms: 1092.8051
type: 'test'
...
# Subtest: model registry keeps original credentials/endpoint intact, rejects stale edits, and forgets only volatile keys on restart
ok 7 - model registry keeps original credentials/endpoint intact, rejects stale edits, and forgets only volatile keys on restart
---
duration_ms: 1416.2692
type: 'test'
...
# Subtest: config import previews exact revisions, keeps secrets volatile, syncs only when enabled and preserves manual edits on conflicts
ok 8 - config import previews exact revisions, keeps secrets volatile, syncs only when enabled and preserves manual edits on conflicts
---
duration_ms: 990.0314
type: 'test'
...
# Subtest: real HTTP pool fails over 429 to another model and never repeats completed tools; attempts seal actual request bytes and Run identity
ok 9 - real HTTP pool fails over 429 to another model and never repeats completed tools; attempts seal actual request bytes and Run identity
---
duration_ms: 3079.9908
type: 'test'
...
# Subtest: pool rotation and weights select members on new Runs; disabled and missing credentials are excluded without requests
ok 10 - pool rotation and weights select members on new Runs; disabled and missing credentials are excluded without requests
---
duration_ms: 5018.7778
type: 'test'
...
# Subtest: 401 blocks the credential revision; updating the key restores availability without leaking error echoes
ok 11 - 401 blocks the credential revision; updating the key restores availability without leaking error echoes
---
duration_ms: 1973.4865
type: 'test'
...
# Subtest: queued request cancellation and in-flight disable release only owned leases and do not dispatch another HTTP request
ok 12 - queued request cancellation and in-flight disable release only owned leases and do not dispatch another HTTP request
---
duration_ms: 963.6111
type: 'test'
...
# Subtest: compiled model-center HTTP API persists public profiles, blocks forged origins, and local operations never invoke a model
ok 13 - compiled model-center HTTP API persists public profiles, blocks forged origins, and local operations never invoke a model
---
duration_ms: 2006.8513
type: 'test'
...
# Subtest: model registry disk failures restore routing, pool profiles and credential references without partial updates
ok 14 - model registry disk failures restore routing, pool profiles and credential references without partial updates
---
duration_ms: 566.9346
type: 'test'
...
# Subtest: model selection and saved configuration commit together and failures preserve the original mode and endpoint
ok 15 - model selection and saved configuration commit together and failures preserve the original mode and endpoint
---
duration_ms: 693.7432
type: 'test'
...
# Subtest: configuration imports roll back the entire batch and retain one usable preview after validation or commit failure
ok 16 - configuration imports roll back the entire batch and retain one usable preview after validation or commit failure
---
duration_ms: 574.4074
type: 'test'
...
# Subtest: claiming the same runtime owner twice preserves its original metadata and every database row
ok 17 - claiming the same runtime owner twice preserves its original metadata and every database row
---
duration_ms: 1095.9592
type: 'test'
...
# Subtest: sessions without controlled-engine history retain their prior export shape and never expose another session record
ok 18 - sessions without controlled-engine history retain their prior export shape and never expose another session record
---
duration_ms: 995.0958
type: 'test'
...
# Subtest: live workers, controlled drivers, or MCP operations refuse an upgrade before changing handlers, managers or database rows
ok 19 - live workers, controlled drivers, or MCP operations refuse an upgrade before changing handlers, managers or database rows
---
duration_ms: 949.7472
type: 'test'
...
# Subtest: an idle upgrade loads the newly introduced engine record map, resets obsolete managers and preserves original rows, sessions and provider objects
ok 20 - an idle upgrade loads the newly introduced engine record map, resets obsolete managers and preserves original rows, sessions and provider objects
---
duration_ms: 1051.4217
type: 'test'
...
# Subtest: an installation failure restores old managers, original record maps, handlers and all existing database rows
ok 21 - an installation failure restores old managers, original record maps, handlers and all existing database rows
---
duration_ms: 1565.5388
type: 'test'
...
# Subtest: the owned controlled driver remains an active session and project even after outer task activity has ended
ok 22 - the owned controlled driver remains an active session and project even after outer task activity has ended
---
duration_ms: 1000.6998
type: 'test'
...
# Subtest: release and upgrade help or invalid arguments have no file or network setup side effects in an independent directory
ok 23 - release and upgrade help or invalid arguments have no file or network setup side effects in an independent directory
---
duration_ms: 990.8517
type: 'test'
...
1..23
# tests 23
# suites 0
# pass 23
# fail 0
# cancelled 0
# skipped 0
# todo 0
# duration_ms 17409.5834
+105
View File
@@ -0,0 +1,105 @@
{
"success": true,
"files": [
{
"source": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-a-accepted.tap",
"path": "docs/desktop-host-a-accepted.tap",
"bytes": 13730,
"sha256": "c0aaa540a682d8c9520e5e7f823370f42e4f1ee3f7e87db899f132f009eb7695",
"previousCommittedSha256": "a738c0a6403ab1636d9626ffd8d1d5d996ae7f22f9a02e14c901835aab75c7dc",
"previousCommittedBytes": 13426,
"byteNormalizationCorrection": true
},
{
"source": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-a-first-failures.txt",
"path": "docs/desktop-host-a-first-failures.txt",
"bytes": 8505,
"sha256": "8e6d069ea668e144ef338c9c4b7a56ed0b1d9d7f91451740736ae18ec5306699",
"previousCommittedSha256": "8e6d069ea668e144ef338c9c4b7a56ed0b1d9d7f91451740736ae18ec5306699",
"previousCommittedBytes": 8505,
"byteNormalizationCorrection": false
},
{
"source": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-a-second-failure.tap",
"path": "docs/desktop-host-a-acl-failure.tap",
"bytes": 14870,
"sha256": "8cdadc5290ff2fbe2dc4b928847846a741df0f9d8851348127d42e6d0ee2a7bf",
"previousCommittedSha256": "5534e9c1bc23690c93dfc4eb2b785f8ca74c2923bca7e9159e20f33db3cbb205",
"previousCommittedBytes": 14551,
"byteNormalizationCorrection": true
},
{
"source": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-a-repair.tap",
"path": "docs/desktop-host-a-repair.tap",
"bytes": 6954,
"sha256": "4e650e5e89b4e403bdbc1d668a6b05a98d65b959c181377a9b1bbda58f940178",
"previousCommittedSha256": "e0c28116f1a935eb60a6c98e9a611e7a8877a87d35e286c925b7f20a09fb09fa",
"previousCommittedBytes": 6806,
"byteNormalizationCorrection": true
},
{
"source": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-a-real-final.log",
"path": "docs/desktop-host-a-actual.txt",
"bytes": 450,
"sha256": "2ebbc135eecd1c733f8c6ac885b07d1e0845d1ffceb24632469872b72f017457",
"previousCommittedSha256": "53e80e49388f47b3b9f0b304f71876dc551239180236bfd198188fafb911180b",
"previousCommittedBytes": 441,
"byteNormalizationCorrection": true
},
{
"source": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-b-tests-first.tap",
"path": "docs/desktop-credentials-b-tests.tap",
"bytes": 6371,
"sha256": "58004a6eaeb141d132bcb963e2fff8edee3dceacb393bdf7ab1037b280a074f3",
"previousCommittedSha256": "1d571a9e06636877a5447275cc0d5609446ddbdc46f1f2efd5807dd583a4fdf0",
"previousCommittedBytes": 6235,
"byteNormalizationCorrection": true
},
{
"source": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-b-ui-final.txt",
"path": "docs/desktop-credentials-b-actual.txt",
"bytes": 465,
"sha256": "c78fd8cc8b73ef415bc87b9d3ac7720e23513537d2a8aa8302a545f38a4886d4",
"previousCommittedSha256": "f55cf1123c1d5ff05def023dac28238959c48508e73cece8ce00ce0d3523e10d",
"previousCommittedBytes": 458,
"byteNormalizationCorrection": true
},
{
"source": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-b-ui-second-run.txt",
"path": "docs/desktop-credentials-b-prelayout.txt",
"bytes": 465,
"sha256": "c78fd8cc8b73ef415bc87b9d3ac7720e23513537d2a8aa8302a545f38a4886d4",
"previousCommittedSha256": "f55cf1123c1d5ff05def023dac28238959c48508e73cece8ce00ce0d3523e10d",
"previousCommittedBytes": 458,
"byteNormalizationCorrection": true
},
{
"source": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-host-a0f264b1-949f-413a-9be1-c7d29823eb40\\result.json",
"path": "docs/desktop-host-a-actual.json",
"bytes": 1576,
"sha256": "591b52274540876513c236f0beb73371fdbf6b5abafc36feaed1f63f385c49c6",
"previousCommittedSha256": "591b52274540876513c236f0beb73371fdbf6b5abafc36feaed1f63f385c49c6",
"previousCommittedBytes": 1576,
"byteNormalizationCorrection": false
},
{
"source": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4083fa57-495c-4f03-a634-38b3af7ac525\\result.json",
"path": "docs/desktop-credentials-b-actual.json",
"bytes": 4993,
"sha256": "4fc932a831b03838a926163a825392c17db98f4cfda03449c9d3063037baf5e6",
"previousCommittedSha256": "37415bc541150e8de73dfcafb433dd1cbf80c6f0bbcdba733270434ad417f401",
"previousCommittedBytes": 4899,
"byteNormalizationCorrection": true
},
{
"source": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-credentials-4cfcf5da-eaaf-472e-b6c0-99266b1e9a3c\\result.json",
"path": "docs/desktop-credentials-b-prelayout.json",
"bytes": 1646,
"sha256": "95a3c6ce02971fe8e01a1b109b4b116caf53411998884c73e3f4e3a15bd65876",
"previousCommittedSha256": "eff2d8c53cf5e6eb08216dad3e9e2f4af611c12ed108e71c85d15e4f92b0598b",
"previousCommittedBytes": 1610,
"byteNormalizationCorrection": true
}
],
"scope": "Original A/B raw evidence bytes retained. Explicit -text attributes repair earlier Git line-ending normalization only; historical commits/reports remain unchanged."
}
+14
View File
@@ -0,0 +1,14 @@
> local-agent-workbench@1.15.0 build
> tsc --noEmit && vite build && node scripts/build-server.mjs
You are using Node.js 16.13.1. Vite requires Node.js version 20.19+ or 22.12+. Please upgrade your Node.js version.
file:///D:/Ai-project/REA%E9%80%86%E5%90%91/Agent-Workbench/node_modules/vite/dist/node/chunks/config.js:8
import fsp, { constants } from "node:fs/promises";
^^^^^^^^^
SyntaxError: The requested module 'node:fs/promises' does not provide an export named 'constants'
at ModuleJob._instantiate (node:internal/modules/esm/module_job:124:21)
at async ModuleJob.run (node:internal/modules/esm/module_job:181:5)
at async Promise.all (index 0)
at async ESMLoader.import (node:internal/modules/esm/loader:281:24)
at async CAC.<anonymous> (file:///D:/Ai-project/REA%E9%80%86%E5%90%91/Agent-Workbench/node_modules/vite/dist/node/cli.js:626:28)
File diff suppressed because it is too large Load Diff
+14
View File
@@ -0,0 +1,14 @@
> local-agent-workbench@1.15.0 build
> tsc --noEmit && vite build && node scripts/build-server.mjs
vite v7.3.7 building client environment for production...
transforming...
✓ 1718 modules transformed.
rendering chunks...
computing gzip size...
dist-client/index.html  0.50 kB │ gzip: 0.37 kB
dist-client/assets/index-CxHXL0MO.css 130.90 kB │ gzip: 25.10 kB
dist-client/assets/index-DluFw6fs.js 496.30 kB │ gzip: 146.12 kB
✓ built in 1.44s
Server bundle built.
File diff suppressed because it is too large Load Diff
+42
View File
@@ -0,0 +1,42 @@
{
"success": true,
"files": [
{
"source": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-host-f-build-first.log",
"path": "docs/desktop-host-f-build-first-log.txt",
"bytes": 894,
"sha256": "0ebbeb738fa7fe2ccdbda84b098e2ddf376cd1738f891e8f52861c99e9ef9f3a"
},
{
"source": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-host-f-build-first-validation.json",
"path": "docs/desktop-host-f-build-first-result.json",
"bytes": 42804,
"sha256": "79c82a8580f617a0e32ef95d22b696beaaf070261c801a05d043af15717c547f"
},
{
"source": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-host-f-build-second-log.txt",
"path": "docs/desktop-host-f-build-second-log.txt",
"bytes": 685,
"sha256": "8c346201fb678d538cf68be99cc3216271f9a64369640ee52e20994cfc75d388"
},
{
"source": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-host-f-build-second-result.json",
"path": "docs/desktop-host-f-build-second-result.json",
"bytes": 57256,
"sha256": "fa7cdcdc96ebec54ddc4f496085b147c1b7680634c25e5305e3887520c4a4985"
},
{
"source": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-host-f-related-first-log.txt",
"path": "docs/desktop-host-f-related-first-log.txt",
"bytes": 4791,
"sha256": "682d35ca4e4c2622b92dc7c888adfd06a76607e69787f5c74a68756cf2c716e6"
},
{
"source": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\desktop-host-f-related-first-result.json",
"path": "docs/desktop-host-f-related-first-result.json",
"bytes": 57451,
"sha256": "0af693170de720831a8a9d41f74ad333c65477890a862215491be12722f1f0e5"
}
],
"scope": "Original build environment failure and subsequent fixed-Node build plus 17 related tests. Not a complete release acceptance."
}
+112
View File
@@ -0,0 +1,112 @@
TAP version 13
# Subtest: a rebuilt source archive binds actual generated client assets and retains its original running identity after disk changes
ok 1 - a rebuilt source archive binds actual generated client assets and retains its original running identity after disk changes
---
duration_ms: 2148.6442
type: 'test'
...
# Subtest: host challenge proof rejects a stale challenge, changed identity and changed signature
ok 2 - host challenge proof rejects a stale challenge, changed identity and changed signature
---
duration_ms: 0.8172
type: 'test'
...
# Subtest: stop confirmation binds the original instance, creation and loaded runtime
ok 3 - stop confirmation binds the original instance, creation and loaded runtime
---
duration_ms: 0.1437
type: 'test'
...
# Subtest: state paths reject escaping, linked ancestors, ambiguous hardlinks and damaged JSON
ok 4 - state paths reject escaping, linked ancestors, ambiguous hardlinks and damaged JSON
---
duration_ms: 1148.0079
type: 'test'
...
# Subtest: failed registration publishes no initialized memory and can complete after its actual inputs are supplied
ok 5 - failed registration publishes no initialized memory and can complete after its actual inputs are supplied
---
duration_ms: 2313.9959
type: 'test'
...
# Subtest: host shutdown cancels an un-signalled HTTP model request and refuses later requests until recovered
ok 6 - host shutdown cancels an un-signalled HTTP model request and refuses later requests until recovered
---
duration_ms: 1.4393
type: 'test'
...
# Subtest: host shutdown cancels model-pool acquisition and releases queue ownership without inventing attempts
ok 7 - host shutdown cancels model-pool acquisition and releases queue ownership without inventing attempts
---
duration_ms: 748.7107
type: 'test'
...
# Subtest: full candidate validation returns the exact loaded-input host identity and rejects undeclared or altered files
ok 8 - full candidate validation returns the exact loaded-input host identity and rejects undeclared or altered files
---
duration_ms: 1281.3903
type: 'test'
...
# Subtest: installation requires separate absolute program/data directories and a candidate in its own versions namespace
ok 9 - installation requires separate absolute program/data directories and a candidate in its own versions namespace
---
duration_ms: 818.2452
type: 'test'
...
# Subtest: staging retains the original data and pointer without querying or restarting a host
ok 10 - staging retains the original data and pointer without querying or restarting a host
---
duration_ms: 2187.3817
type: 'test'
...
# Subtest: a post-pointer failure restores every original pointer byte and persistent row; a first-install failure removes only its published pointer
ok 11 - a post-pointer failure restores every original pointer byte and persistent row; a first-install failure removes only its published pointer
---
duration_ms: 4270.6777
type: 'test'
...
# Subtest: changed active manifest and concurrent pointer changes are rejected without overwriting the unknown record
ok 12 - changed active manifest and concurrent pointer changes are rejected without overwriting the unknown record
---
duration_ms: 1802.8301
type: 'test'
...
# Subtest: an unregistered non-HTTP listener blocks preflight even when service-info is absent
ok 13 - an unregistered non-HTTP listener blocks preflight even when service-info is absent
---
duration_ms: 926.0963
type: 'test'
...
# Subtest: uninstall archives its original pointer and preserves every data row/artifact instead of recursively deleting data
ok 14 - uninstall archives its original pointer and preserves every data row/artifact instead of recursively deleting data
---
duration_ms: 2718.4972
type: 'test'
...
# Subtest: preflight rejects an unrelated program root or changed native bootstrap while preserving its original bytes
ok 15 - preflight rejects an unrelated program root or changed native bootstrap while preserving its original bytes
---
duration_ms: 844.1368
type: 'test'
...
# Subtest: a changed owned bootstrap follows its candidate and a post-pointer fault restores both original bytes
ok 16 - a changed owned bootstrap follows its candidate and a post-pointer fault restores both original bytes
---
duration_ms: 4417.8842
type: 'test'
...
# Subtest: an external bootstrap change after preflight is preserved and never overwritten as an update or rollback
ok 17 - an external bootstrap change after preflight is preserved and never overwritten as an update or rollback
---
duration_ms: 2819.1904
type: 'test'
...
1..17
# tests 17
# suites 0
# pass 17
# fail 0
# cancelled 0
# skipped 0
# todo 0
# duration_ms 22180.5487
File diff suppressed because it is too large Load Diff
+5
View File
@@ -638,3 +638,8 @@ C实际目录选择完成:sky元数据焦点/重复索引与原生光标不一
Electron验收器不能假定后台SDK事件停止:seq合法前进时,原持久事件字节、原seq的HTTP补发、同Task/Run/审批/完成工具行仍需核对。父成果按钮导航到改动页后,还需要实际“审查变更集”动作;隐藏原生窗口的截图可能等待绘制。报告下载直接用实际DownloadItem并核对原SHA,QA仅固定自有保存路径,Windows保存框未测。E完整工程/F发行仍需完成。
E339输入/16节点与原生九组已经完整验收;原失败和67份证据/QA helper单文件差异保持。手机仅有1.16计划,F仍须完整发行与正式部署。
## 1.15 F发行前核对
A–E已提交并保留原证据。F源码包必须对实际重新生成的客户端资产建立宿主身份、使用随包Electron和ignore-scripts离线安装,并核对全部dist目录与示例/研究/基准输入。便携验收跟随D的示例用户数据初始化,不再期待示例写在不可变程序目录;运行后重新核对所有程序原字节。首个build被系统Node16抢占,原结果保留;固定Node24.22.1后构建及17相关检查通过。A/B原工作区/原QA报告语义一致,但九份旧Git日志/JSON被自动换行,已用明确-text属性与原字节修正当前索引,历史commit不改写。15份Markdown/294链接及A–E原证据/索引核对通过;源码整轮/正式发行继续执行。几次猜测源码/报告路径的读取失败无产品副作用,随后按rg实际文件定位。
+2 -2
View File
@@ -1,12 +1,12 @@
{
"name": "local-agent-workbench",
"version": "1.14.0",
"version": "1.15.0",
"lockfileVersion": 3,
"requires": true,
"packages": {
"": {
"name": "local-agent-workbench",
"version": "1.14.0",
"version": "1.15.0",
"dependencies": {
"@babel/parser": "7.29.11",
"diff": "^8.0.2",
+6 -3
View File
@@ -1,6 +1,6 @@
{
"name": "local-agent-workbench",
"version": "1.14.0",
"version": "1.15.0",
"private": true,
"type": "module",
"description": "A local coding workbench with explicit tool execution, reviewable changes and recoverable sessions.",
@@ -8,7 +8,7 @@
"dev": "vite --host 127.0.0.1",
"build": "tsc --noEmit && vite build && node scripts/build-server.mjs",
"start": "node dist-server/server.mjs",
"desktop": "electron dist-desktop/main.cjs",
"desktop": "powershell.exe -NoProfile -ExecutionPolicy Bypass -File scripts/Start-Desktop.ps1",
"verify:desktop": "node scripts/verify-desktop-window.mjs",
"test": "node --test --test-concurrency=2 tests/*.test.mjs",
"verify:ui": "node scripts/verify-ui.mjs",
@@ -34,7 +34,10 @@
"verify:controlled": "node scripts/verify-controlled-ui.mjs",
"verify:controlled-projects": "node scripts/verify-controlled-projects.mjs",
"verify:purge-ui": "node scripts/verify-purge-ui.mjs",
"verify:purge-comparison": "node scripts/verify-controlled-projects.mjs --purge-comparison"
"verify:purge-comparison": "node scripts/verify-controlled-projects.mjs --purge-comparison",
"verify:desktop-projects": "node scripts/verify-controlled-projects.mjs --desktop",
"verify:desktop-lifecycle": "node scripts/verify-controlled-ui.mjs --desktop",
"verify:desktop-install": "node scripts/verify-desktop-install.mjs"
},
"dependencies": {
"@babel/parser": "7.29.11",
+5
View File
@@ -670,3 +670,8 @@ E首轮环境根误用、精确seq误判、父审查漏按钮、隐藏窗口截
E新整轮三个工程16节点通过,339输入/全部构建保持。父Node8/8、React9/9、Python10/10和实际构建/编译、父审查/Git通过,7个原生Run完整退出;各一次失败修复/两次审批和7/7/9次自有HTTP记账保持。12张真实Electron1440/390截图与原DOM封存;代表性宽窄截图人工核对。原生UI九组/实际DownloadItem/停止/unknown通过,67份原失败/成功证据核对;其后唯一QA helper布局动作差异单独登记,F最终1.15复验。D提交f7a5969,正式仍1.14/PID23172/hasKey=true/16项目/20可见会话,无活动任务;没有读取Key。
F下一步同步1.15版本/独立证据前缀,离线源码运行依赖和全部dist目录,完整源码/历史UI/桌面工程/安装与正式状态保留升级。1.16配对伴随端计划已写,只有1.15 F全部完成后才实施;本轮不建立新代理、goal或自动化。
## 2026-10-31 F发行接续
A–E提交与原证据已核对,HEAD b77783b。统一1.15版本/独立证据路径,离线源码支持随包Electron/ignore-scripts,源码重建资产进入实际宿主身份;全部dist构建范围扩展。首个build误用系统Node16,原日志保存;QA子进程固定Node24.22.1后构建成功,17相关宿主/窗口/安装/旧升级检查通过。正式仍1.14/PID23172;接着冻结/完整历史UI/独立源码/最终发行安装与保留状态部署,PLAN-1.16仅记录。
+3 -1
View File
@@ -7,7 +7,9 @@ const digest = file => createHash('sha256').update(fs.readFileSync(file)).digest
export function planDesktopBundle(root) {
const meta = JSON.parse(fs.readFileSync(path.join(root, 'node_modules/electron/package.json'), 'utf8'));
if (meta.version !== '45.2.0') throw new Error('Desktop packaging requires the fixed Electron 45.2.0 runtime.');
const directory = path.join(root, 'node_modules/electron/dist'), executable = path.join(directory, 'electron.exe');
const installed = path.join(root, 'node_modules/electron/dist');
const directory = fs.existsSync(path.join(installed, 'electron.exe')) ? installed : path.join(root, 'runtime/electron');
const executable = path.join(directory, 'electron.exe');
if (!fs.existsSync(executable) || digest(executable) !== '010ea51bfb528c8143513439aa93fb40754403daa2023f9decd4517da71706ab') throw new Error('The actual Windows Electron runtime differs from the fixed binary.');
for (const file of ['LICENSE', 'LICENSES.chromium.html']) if (!fs.existsSync(path.join(directory, file))) throw new Error('Electron runtime license missing.');
return { directory, version: meta.version, sha256: digest(executable), bytes: fs.statSync(executable).size };
+1
View File
@@ -7,6 +7,7 @@ const prefixes = Object.freeze({
'1.12.1': 'milestone-audit',
'1.13.0': 'workers',
'1.14.0': 'controlled-engine',
'1.15.0': 'desktop-host',
});
export function releaseEvidencePrefix(version) {
const prefix = prefixes[version];
+11 -11
View File
@@ -7,28 +7,28 @@ import { gitEnvironment } from '../src/core/git-process.mjs';
import { randomUUID } from 'node:crypto';
if (process.argv.slice(2).some(value => value !== '--help')) throw new Error('Unsupported argument; use --help.');
if (process.argv.includes('--help')) { console.log('node scripts/verify-all-ui.mjs - run serial desktop/390px checks in independent local hosts; use explicit QA Python/Harness/Purge source paths for 1.14.'); process.exit(0); }
if (process.argv.includes('--help')) { console.log('node scripts/verify-all-ui.mjs - run serial desktop/390px checks in independent local hosts; use explicit QA Python/Harness/Purge source paths for 1.15.'); process.exit(0); }
const version = JSON.parse(fs.readFileSync('package.json', 'utf8')).version;
assert.ok(['1.6.0', '1.6.1', '1.6.2', '1.7.0', '1.8.0', '1.9.0', '1.10.0', '1.10.1', '1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0'].includes(version), 'Review evidence paths when introducing the next release.');
assert.ok(['1.6.0', '1.6.1', '1.6.2', '1.7.0', '1.8.0', '1.9.0', '1.10.0', '1.10.1', '1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0', '1.15.0'].includes(version), 'Review evidence paths when introducing the next release.');
const prefix = releaseEvidencePrefix(version), resultFile = `.qa/${prefix}-all-ui-validation.json`, results = [];
const qaEnvironment = Object.fromEntries(['AGENT_WORKBENCH_TEST_PYTHON', 'AGENT_WORKBENCH_DEEPSEEK_ROOT', 'AGENT_WORKBENCH_PURGE_178_ROOT', 'AGENT_WORKBENCH_PURGE_181_ROOT']
.filter(name => process.env[name]).map(name => [name, process.env[name]]));
for (const [name, value] of Object.entries(qaEnvironment)) assert.ok(path.isAbsolute(value) && fs.existsSync(value), 'Use an existing absolute QA runtime/source path: ' + name);
const cases = [
...(version === '1.14.0' ? [['verify-controlled-ui.mjs', null], ['verify-purge-ui.mjs', null]] : []),
...(['1.13.0', '1.14.0'].includes(version) ? [['verify-workers-ui.mjs', null]] : []),
...(['1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0'].includes(version) ? [['verify-task-plans-ui.mjs', `docs/${prefix}-task-ui-validation.json`]] : []),
...(['1.10.0', '1.10.1', '1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0'].includes(version) ? [['verify-browser-ui.mjs', `docs/${prefix}-browser-ui-validation.json`]] : []),
...(['1.14.0', '1.15.0'].includes(version) ? [['verify-controlled-ui.mjs', null], ['verify-purge-ui.mjs', null]] : []),
...(['1.13.0', '1.14.0', '1.15.0'].includes(version) ? [['verify-workers-ui.mjs', null]] : []),
...(['1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0', '1.15.0'].includes(version) ? [['verify-task-plans-ui.mjs', `docs/${prefix}-task-ui-validation.json`]] : []),
...(['1.10.0', '1.10.1', '1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0', '1.15.0'].includes(version) ? [['verify-browser-ui.mjs', `docs/${prefix}-browser-ui-validation.json`]] : []),
['verify-ui.mjs', '.qa/ui-validation.json'], ['verify-diagnostics-ui.mjs', '.qa/diagnostics-ui-validation.json'],
['verify-execution-ui.mjs', '.qa/execution-ui-validation.json'], ['verify-delivery-ui.mjs', '.qa/delivery-ui-validation.json'],
['verify-index-ui.mjs', '.qa/index-ui-result.json'], ['verify-change-sets-ui.mjs', `docs/${prefix}-batch-ui-validation.json`],
['verify-context-ui.mjs', `.qa/${version === '1.6.0' ? 'workspaces-context' : prefix}-ui-validation.json`],
['verify-workspaces-ui.mjs', `docs/${prefix === 'workspaces' ? prefix : prefix + '-workspaces'}-ui-validation.json`],
...(['1.7.0', '1.8.0', '1.9.0', '1.10.0', '1.10.1', '1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0'].includes(version) ? [['verify-python-ui.mjs', `docs/${prefix}-python-ui-validation.json`]] : []),
...(['1.7.0', '1.8.0', '1.9.0', '1.10.0', '1.10.1', '1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0'].includes(version) ? [['verify-react-delivery.mjs', version === '1.7.0' ? 'docs/react-engine-delivery-validation.json' : `docs/${prefix}-react-delivery-validation.json`]] : []),
...(['1.8.0', '1.9.0', '1.10.0', '1.10.1', '1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0'].includes(version) ? [['verify-research-ui.mjs', `docs/${prefix}-research-ui-validation.json`]] : []),
...(['1.9.0', '1.10.0', '1.10.1', '1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0'].includes(version) ? [['verify-engines-ui.mjs', version === '1.9.0' ? 'docs/engines-ui-validation.json' : `docs/${prefix}-engines-ui-validation.json`]] : []),
...(['1.12.0', '1.12.1', '1.13.0', '1.14.0'].includes(version) ? ['model-center', 'project-memory', 'task-checkpoints', 'skills', 'mcp', 'mcp-recovery'].map(name => [`verify-${name}-ui.mjs`, null]) : []),
...(['1.7.0', '1.8.0', '1.9.0', '1.10.0', '1.10.1', '1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0', '1.15.0'].includes(version) ? [['verify-python-ui.mjs', `docs/${prefix}-python-ui-validation.json`]] : []),
...(['1.7.0', '1.8.0', '1.9.0', '1.10.0', '1.10.1', '1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0', '1.15.0'].includes(version) ? [['verify-react-delivery.mjs', version === '1.7.0' ? 'docs/react-engine-delivery-validation.json' : `docs/${prefix}-react-delivery-validation.json`]] : []),
...(['1.8.0', '1.9.0', '1.10.0', '1.10.1', '1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0', '1.15.0'].includes(version) ? [['verify-research-ui.mjs', `docs/${prefix}-research-ui-validation.json`]] : []),
...(['1.9.0', '1.10.0', '1.10.1', '1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0', '1.15.0'].includes(version) ? [['verify-engines-ui.mjs', version === '1.9.0' ? 'docs/engines-ui-validation.json' : `docs/${prefix}-engines-ui-validation.json`]] : []),
...(['1.12.0', '1.12.1', '1.13.0', '1.14.0', '1.15.0'].includes(version) ? ['model-center', 'project-memory', 'task-checkpoints', 'skills', 'mcp', 'mcp-recovery'].map(name => [`verify-${name}-ui.mjs`, null]) : []),
];
fs.mkdirSync('.qa', { recursive: true });
const externalHost = new Set(['verify-ui.mjs', 'verify-diagnostics-ui.mjs', 'verify-execution-ui.mjs', 'verify-index-ui.mjs']);
+1 -1
View File
@@ -34,7 +34,7 @@ child.stdout.on('data', value => { output += value; for (const line of output.sp
child.stderr.on('data', value => { errors += value; });
const wait = async predicate => { const deadline = Date.now() + 15000; while (!await predicate()) { if (Date.now() > deadline) throw new Error('UI fixture timed out: ' + errors); await new Promise(resolve => setTimeout(resolve, 40)); } };
await wait(() => url); const bootstrap = await (await fetch(url + '/api/bootstrap')).json();
assert.ok(['1.4.0', '1.5.0', '1.5.1', '1.6.0', '1.6.1', '1.6.2', '1.7.0', '1.8.0', '1.9.0', '1.10.0', '1.10.1', '1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0'].includes(bootstrap.version)); assert.equal(bootstrap.config.mode, 'demo'); assert.equal(bootstrap.config.hasKey, false);
assert.ok(['1.4.0', '1.5.0', '1.5.1', '1.6.0', '1.6.1', '1.6.2', '1.7.0', '1.8.0', '1.9.0', '1.10.0', '1.10.1', '1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0', '1.15.0'].includes(bootstrap.version)); assert.equal(bootstrap.config.mode, 'demo'); assert.equal(bootstrap.config.hasKey, false);
const proofPrefix = releaseEvidencePrefix(bootstrap.version) + '-batch';
const screenshotPrefix = bootstrap.version === '1.4.0' ? '1.4-batch' : proofPrefix;
const api = async (route, input) => { const response = await fetch(url + '/api' + route, { method: input === undefined ? 'GET' : 'POST', headers: { 'Content-Type': 'application/json', 'X-Workbench-Token': bootstrap.token, Connection: 'close' },
+1 -1
View File
@@ -77,7 +77,7 @@ child.stderr.on('data', bytes => { errors += bytes.toString('utf8'); });
const until = async predicate => { const deadline = Date.now() + 20000; while (!await predicate()) { if (Date.now() > deadline) throw new Error('Context UI fixture timeout: ' + errors); await new Promise(resolve => setTimeout(resolve, 40)); } };
await until(() => url || child.exitCode !== null); assert.ok(url, errors);
const bootstrap = await (await fetch(url + '/api/bootstrap', { headers: { Connection: 'close' } })).json();
assert.ok(['1.5.0', '1.5.1', '1.6.0', '1.6.1', '1.6.2', '1.7.0', '1.8.0', '1.9.0', '1.10.0', '1.10.1', '1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0'].includes(bootstrap.version)); assert.equal(bootstrap.config.mode, 'demo'); assert.equal(bootstrap.config.hasKey, false);
assert.ok(['1.5.0', '1.5.1', '1.6.0', '1.6.1', '1.6.2', '1.7.0', '1.8.0', '1.9.0', '1.10.0', '1.10.1', '1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0', '1.15.0'].includes(bootstrap.version)); assert.equal(bootstrap.config.mode, 'demo'); assert.equal(bootstrap.config.hasKey, false);
const api = async route => { const response = await fetch(url + '/api' + route, { headers: { 'X-Workbench-Token': bootstrap.token, Connection: 'close' } }); const value = await response.json(); assert.ok(response.ok, value.error); return value; };
const proofPrefix = bootstrap.version === '1.6.0' ? 'workspaces-context' : releaseEvidencePrefix(bootstrap.version);
const browser = await chromium.launch({ channel: 'msedge', headless: true }), page = await browser.newPage({ viewport: { width: 1440, height: 1000 } });
+3 -3
View File
@@ -17,7 +17,7 @@ import { desktopProjectQa } from './desktop-project-qa.mjs';
const args = process.argv.slice(2);
assert.ok(args.every(value => ['--help', '--purge-comparison', '--desktop'].includes(value)), 'Unknown option; use --help.');
if (args.includes('--help')) {
console.log('node scripts/verify-controlled-projects.mjs [--purge-comparison] — actual pinned Harness and owned HTTP model fixture; default verifies Node/React/Python with approvals, Edge and Git. Comparison verifies eight independent Node projects: two Purge versions, two repetitions, enabled/disabled. Set AGENT_WORKBENCH_DEEPSEEK_ROOT; default also requires AGENT_WORKBENCH_TEST_PYTHON, comparison requires AGENT_WORKBENCH_PURGE_178_ROOT / AGENT_WORKBENCH_PURGE_181_ROOT. No configured-model requests.'); process.exit(0);
console.log('node scripts/verify-controlled-projects.mjs [--purge-comparison | --desktop] — actual pinned Harness and owned HTTP model fixture; default verifies Node/React/Python with approvals, Edge and Git. Comparison verifies eight independent Node projects: two Purge versions, two repetitions, enabled/disabled. Set AGENT_WORKBENCH_DEEPSEEK_ROOT; default also requires AGENT_WORKBENCH_TEST_PYTHON, comparison requires AGENT_WORKBENCH_PURGE_178_ROOT / AGENT_WORKBENCH_PURGE_181_ROOT. No configured-model requests.'); process.exit(0);
}
const comparisonMode = args.includes('--purge-comparison'), desktopMode = args.includes('--desktop');
assert.ok(!comparisonMode || !desktopMode, 'Desktop engineering and prompt comparison are separate acceptance runs.');
@@ -252,7 +252,7 @@ try {
assert.equal(results.length, 8); assert.equal(comparisons.length, 4);
} else for (const kind of ['node', 'react', 'python']) results.push(await runSample(kind));
assert.deepEqual(inputs(), frozen);
report = { success: true, targetVersion: '1.14.0', testedVersion: version, directory, installation, cases, results, comparisons, sources, inputs: frozen, realModelRequests: 0,
report = { success: true, targetVersion: version, testedVersion: version, directory, installation, cases, results, comparisons, sources, inputs: frozen, realModelRequests: 0,
scope: 'Pinned actual Harness and workbench execution with an owned HTTP model fixture. Checks, approvals, failures, parent merges, Edge and Git are real. Model reasoning and Purge benefit remain separate.' };
} catch (error) { report = { success: false, targetVersion: '1.14.0', testedVersion: version, directory, installation, cases, results, inputs: frozen, error: error.stack, realModelRequests: 0 }; process.exitCode = 1; }
} catch (error) { report = { success: false, targetVersion: version, testedVersion: version, directory, installation, cases, results, inputs: frozen, error: error.stack, realModelRequests: 0 }; process.exitCode = 1; }
finally { fs.writeFileSync(path.join(directory, 'result.json'), JSON.stringify(report, null, 2)); fs.writeFileSync('.qa/' + (desktopMode ? 'desktop-projects' : comparisonMode ? 'controlled-purge' : 'controlled-projects') + '-location.json', JSON.stringify({ directory, result: path.join(directory, 'result.json') })); console.log('EVIDENCE ' + directory); console.log(JSON.stringify({ success: report.success, cases: cases.length, directory, realModelRequests: 0, error: report.error ?? null })); }
+3 -1
View File
@@ -21,7 +21,9 @@ async function until(predicate, timeout = 45000) {
export async function verifyControlledApi({ api, url, token, directory, appRoot, contained = false }) {
const origin = new URL(url);
assert.equal(origin.hostname, '127.0.0.1'); assert.notEqual(origin.port, '4318');
const bootstrap = await api('/bootstrap'); assert.equal(bootstrap.version, '1.14.0');
const expectedVersion = JSON.parse(fs.readFileSync(path.join(appRoot, fs.existsSync(path.join(appRoot, 'MANIFEST.json')) ? 'MANIFEST.json' : 'package.json'), 'utf8')).version;
assert.ok(['1.14.0', '1.15.0'].includes(expectedVersion));
const bootstrap = await api('/bootstrap'); assert.equal(bootstrap.version, expectedVersion);
assert.equal(bootstrap.config.mode, 'demo'); assert.equal(bootstrap.config.hasKey, false);
const root = path.join(directory, 'compiled-controlled-node'), hooks = path.join(directory, 'disabled-hooks');
fs.mkdirSync(root); fs.mkdirSync(hooks);
+2 -2
View File
@@ -134,10 +134,10 @@ try {
assert.equal(stoppedJob.status, 'resolved_external'); assert.deepEqual(stoppedJob.originalOutcome, originalOutcome); assert.equal(f.store.artifactStore.read(f.store.core.artifact.get(stoppedJob.reportArtifactId)), originalReport); assert.deepEqual(stopped.usage, previousUsage); assert.deepEqual(ownerBytes(f), original);
mark('390px recovery requires reading the owned observation before explicit review and preserves original outcome/report and cumulative budget');
assert.deepEqual(inputs(), frozen); assert.deepEqual(pageErrors, []); assert.deepEqual(requestFailures, []);
result = { passed: true, targetVersion: '1.14.0', testedVersion: version, assertions, screenshots, pageErrors, requestFailures, expectedRejections, installation,
result = { passed: true, targetVersion: version, testedVersion: version, assertions, screenshots, pageErrors, requestFailures, expectedRejections, installation,
engineRuns: [...f.store.core.engine_run.values()], groups: [group.id, stopped.id], inputs: frozen, realModelRequests: 0, ...(desktop ? { desktopEvents: desktop.events } : {}),
scope: 'Actual native Harness and ' + (desktop ? 'Electron with fixed main/preload and own in-process Node host' : 'Edge') + ' with an owned predefined in-memory model fixture. Unknown recovery is a declared fixture transition; actual host-crash behavior is independently tested.' };
} catch (error) { result = { passed: false, targetVersion: '1.14.0', testedVersion: version, error: error.stack, assertions, screenshots, pageErrors, requestFailures, realModelRequests: 0 }; await screenshot('controlled-ui-failure').catch(() => {}); throw error; }
} catch (error) { result = { passed: false, targetVersion: version, testedVersion: version, error: error.stack, assertions, screenshots, pageErrors, requestFailures, realModelRequests: 0 }; await screenshot('controlled-ui-failure').catch(() => {}); throw error; }
finally {
releaseHeld?.(); await browser?.close(); await desktop?.close(); await host.runtime.stopAll(); f.store.storage.database.exec('PRAGMA wal_checkpoint(TRUNCATE)');
fs.cpSync(f.root, path.join(directory, 'original-fixture'), { recursive: true });
+5 -3
View File
@@ -9,6 +9,7 @@ import { _electron } from 'playwright';
import { HostController } from '../src/core/desktop/host-controller.mjs';
import { observeWindowsProcess } from '../src/core/desktop/process-identity.mjs';
import { projectEnvironment } from '../src/core/process-environment.mjs';
import { desktopExecutableIdentity } from '../src/core/desktop/app-installation.mjs';
if (process.argv.includes('--help')) { console.log('node scripts/verify-desktop-window.mjs [--native-picker] — isolated actual Electron runtime; default uses explicitly identified native-dialog/shell fixtures, --native-picker waits for actual Windows UI selection. No existing credentials or configured model.'); process.exit(0); }
if (process.argv.slice(2).some(arg => arg !== '--native-picker')) throw new Error('Use --help.');
@@ -28,8 +29,9 @@ const completedTools = id => {
try { return db.prepare("SELECT payload FROM records WHERE kind='tool' AND session_id=? ORDER BY id").all(id).map(row => JSON.parse(row.payload)).filter(value => value.state === 'completed'); } finally { db.close(); }
};
async function launch() {
const current = await _electron.launch({ executablePath: path.join(appRoot, 'node_modules/electron/dist/electron.exe'),
args: [path.join(appRoot, 'dist-desktop/main.cjs'), '--app-root', appRoot, '--data-root', dataRoot, '--node', process.execPath, '--port', String(port), ...(native ? [] : ['--hidden-window'])],
const executable = desktopExecutableIdentity(appRoot).path, packaged = executable.toLowerCase().endsWith('agentworkbench.exe');
const current = await _electron.launch({ executablePath: executable,
args: [...(packaged ? [] : [path.join(appRoot, 'dist-desktop/main.cjs')]), '--app-root', appRoot, '--data-root', dataRoot, '--node', process.execPath, '--port', String(port), ...(native ? [] : ['--hidden-window'])],
cwd: appRoot, env: projectEnvironment(), timeout: 60000 });
const window = await current.firstWindow();
window.on('pageerror', error => errors.push(error.message)); window.on('requestfailed', request => { if (!expectedNavigation) requests.push({ url: request.url(), failure: request.failure()?.errorText }); });
@@ -37,7 +39,7 @@ async function launch() {
return { application: current, page: window };
}
async function snapshot(name, width) {
const window = await application.browserWindow(page); await window.evaluate((value, width) => value.setContentSize(width, 900), width);
const window = await application.browserWindow(page); await window.evaluate((value, width) => { value.showInactive(); value.setContentSize(width, 900); }, width);
await until(() => page.evaluate(width => innerWidth === width, width));
await page.screenshot({ path: path.join(root, name + '.png') }); fs.writeFileSync(path.join(root, name + '.html'), await page.content());
const overflow = await page.evaluate(() => document.documentElement.scrollWidth > innerWidth);
+1 -1
View File
@@ -102,7 +102,7 @@ try {
await panel().getByLabel('执行方式', { exact: true }).selectOption('deepseek'); await expect(toggle()).not.toBeChecked(); await expect(experiment()).toContainText('默认关闭');
mark('retaining the original Run and starting the next parent goal restores default-off without changing prior strategy, report or usage');
assert.equal(pageErrors.length, 0); assert.equal(requestFailures.length, 0); assert.deepEqual(inputs(), frozen); assert.deepEqual(ownerBytes(f), original);
result = { passed: true, version, targetVersion: '1.14.0', assertions, screenshots, pageErrors, requestFailures, expectedRejections, sources, groupId: group.id,
result = { passed: true, version, targetVersion: version, assertions, screenshots, pageErrors, requestFailures, expectedRejections, sources, groupId: group.id,
engineRun: job, inputs: frozen, realModelRequests: 0, scope: 'Actual Edge/Harness and an owned in-memory model fixture. Source changes affect declared nine-file copies only; no upstream plugin code executes. Real model benefit remains unmeasured.' };
} catch (error) { result = { passed: false, version, assertions, screenshots, pageErrors, requestFailures, error: error.stack, realModelRequests: 0 }; process.exitCode = 1; }
finally {
+17 -7
View File
@@ -34,7 +34,7 @@ for (const name of Object.keys(env)) if (name.toLowerCase() === 'path') delete e
const windowsRoot = process.env.SystemRoot || 'C:\\Windows';
env.PATH = [windowsRoot, path.join(windowsRoot, 'System32'), path.join(windowsRoot, 'System32', 'WindowsPowerShell', 'v1.0')].join(path.delimiter);
let externalGit = null;
if (manifest.version === '1.14.0') {
if (['1.14.0', '1.15.0'].includes(manifest.version)) {
// The controlled project requires external Git. Preserve its observed binary
// directory while keeping system Node/npm unavailable to the packaged host.
const binary = fs.realpathSync.native(execFileSync('where.exe', ['git.exe'], { env: gitEnvironment(), encoding: 'utf8', windowsHide: true, timeout: 10000 }).trim().split(/\r?\n/)[0]);
@@ -88,7 +88,13 @@ try {
console.log((await launch()).trim());
assert.equal(JSON.parse(fs.readFileSync(path.join(data, 'service-info.json'), 'utf8')).pid, firstInfo.pid);
const example = bootstrap.projects.find(project => project.id === bootstrap.exampleProjectId);
assert.ok(example.root.toLowerCase().startsWith(appRoot.toLowerCase() + path.sep));
if (manifest.version === '1.15.0') {
assert.equal(path.resolve(example.root).toLowerCase(), path.join(data, 'projects', 'starter-project').toLowerCase());
const compareTemplate = relative => { for (const entry of fs.readdirSync(path.join(appRoot, 'examples/starter-project', relative), { withFileTypes: true })) {
const next = path.join(relative, entry.name); if (entry.isDirectory()) compareTemplate(next);
else assert.deepEqual(fs.readFileSync(path.join(example.root, next)), fs.readFileSync(path.join(appRoot, 'examples/starter-project', next)), next);
} }; compareTemplate('');
} else assert.ok(example.root.toLowerCase().startsWith(appRoot.toLowerCase() + path.sep));
const session = await api('/sessions', { projectId: example.id });
await api('/sessions/' + session.id + '/messages', { content: '便携包工具与随附 Node 验证' });
let current;
@@ -108,7 +114,7 @@ try {
const checkLog = await api('/artifacts/' + checked.results[0].logArtifactId + '?session=' + session.id); assert.match(checkLog.content, /tests 4/);
assert.ok(fs.existsSync(path.join(appRoot, 'runtime/node_modules/npm/bin/npm-cli.js')));
const actualRea = [];
if (['1.8.0', '1.9.0', '1.10.0', '1.10.1', '1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0'].includes(manifest.version)) {
if (['1.8.0', '1.9.0', '1.10.0', '1.10.1', '1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0', '1.15.0'].includes(manifest.version)) {
const installation = await api('/research/availability'); assert.equal(installation.status, 'available'); assert.equal(installation.version, '6.8.0');
assert.equal(fs.realpathSync.native(installation.packageRoot), fs.realpathSync.native(path.join(appRoot, 'runtime/node_modules/rea-agents')));
assert.equal(fs.realpathSync.native(installation.node.path), fs.realpathSync.native(path.join(appRoot, 'runtime/node.exe')));
@@ -140,7 +146,7 @@ try {
console.log('PASS actual contained REA: ASAR, PE imports and PE resources; zero target execution/model requests');
}
let actualEngine = null;
if (['1.9.0', '1.10.0', '1.10.1', '1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0'].includes(manifest.version) && process.argv.includes('--native-engines')) {
if (['1.9.0', '1.10.0', '1.10.1', '1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0', '1.15.0'].includes(manifest.version) && process.argv.includes('--native-engines')) {
const installation = await api('/engines/deepseek/availability'); assert.equal(installation.status, 'available', installation.message);
assert.equal(fs.realpathSync.native(installation.node.path), fs.realpathSync.native(path.join(appRoot, 'runtime/node.exe')));
const engineRoot = path.join(data, 'engine-project'); fs.mkdirSync(engineRoot); fs.writeFileSync(path.join(engineRoot, 'README.md'), '# Owned portable engine connection\n');
@@ -160,13 +166,17 @@ try {
assert.equal((await api('/research')).length, 4);
const page = await fetch(url); assert.equal(page.status, 200);
assert.match(await page.text(), /Agent Workbench/);
const actualBrowser = ['1.10.0', '1.10.1', '1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0'].includes(manifest.version) ? await verifyBrowserApi({ api, url, token: bootstrap.token, directory: data, appRoot, contained: true }) : null;
const actualControlled = manifest.version === '1.14.0' ? await verifyControlledApi({ api, url, token: bootstrap.token, directory: data, appRoot, contained: true }) : null;
const actualBrowser = ['1.10.0', '1.10.1', '1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0', '1.15.0'].includes(manifest.version) ? await verifyBrowserApi({ api, url, token: bootstrap.token, directory: data, appRoot, contained: true }) : null;
const actualControlled = ['1.14.0', '1.15.0'].includes(manifest.version) ? await verifyControlledApi({ api, url, token: bootstrap.token, directory: data, appRoot, contained: true }) : null;
console.log((await stop()).trim()); started = false;
await until(async () => { try { await fetch(url + '/api/health'); return false; } catch { return true; } });
for (const file of manifest.files) {
const bytes = fs.readFileSync(path.join(appRoot, file.path)); assert.equal(bytes.length, file.bytes);
assert.equal(createHash('sha256').update(bytes).digest('hex'), file.sha256, file.path);
}
const result = { success: true, date: new Date().toISOString(), appRoot, version: manifest.version, nodeVersion: manifest.nodeVersion,
verifiedFiles: manifest.files.length, url, reusedPid: firstInfo.pid, checks: { jobId, status: checked.status, tests: checked.results[0].tests }, actualRea, actualEngine, actualBrowser, actualControlled, externalGit,
steps: ['manifest hashes', 'bundled Node launch', 'same-service reuse', 'packaged example and static page', 'explicit command consent', 'actual bundled Node version', 'actual bundled npm check without system Node/npm', 'check log artifact', ...(actualRea.length ? ['contained REA installation and Node identity', 'actual ASAR/imports/resources', 'original Bundle and target hash binding', 'verified reuse without producer replay'] : []), 'report availability', 'graceful stop'], realModel: false };
steps: ['manifest hashes', 'bundled Node launch', 'same-service reuse', 'packaged example and static page', ...(manifest.version === '1.15.0' ? ['starter project copied to owned user data; original template bytes matched'] : []), 'explicit command consent', 'actual bundled Node version', 'actual bundled npm check without system Node/npm', 'check log artifact', ...(actualRea.length ? ['contained REA installation and Node identity', 'actual ASAR/imports/resources', 'original Bundle and target hash binding', 'verified reuse without producer replay'] : []), 'report availability', 'graceful stop', 'every original packaged file remains unchanged'], realModel: false };
fs.writeFileSync('.qa/release-validation.json', JSON.stringify(result, null, 2));
console.log(JSON.stringify(result, null, 2));
} finally { if (started) { try { await stop(); } catch {} } }
+25 -10
View File
@@ -19,7 +19,8 @@ assert.ok(fs.existsSync(npmCli), 'Use Node with its bundled npm.');
const files = (root, relative = '') => fs.readdirSync(path.join(root, relative), { withFileTypes: true }).sort((a, b) => a.name.localeCompare(b.name))
.flatMap(entry => entry.isDirectory() ? files(root, path.join(relative, entry.name)) : [path.join(relative, entry.name)]);
const digest = file => createHash('sha256').update(fs.readFileSync(file)).digest('hex');
const expected = Object.fromEntries(['dist-client', 'dist-server'].flatMap(folder => files(path.join(workspace, folder)).map(relative => [path.join(folder, relative), digest(path.join(workspace, folder, relative))])));
const buildFolders = ['dist-client', 'dist-server', ...(evidence.version === '1.15.0' ? ['dist-desktop'] : [])];
const expected = Object.fromEntries(buildFolders.flatMap(folder => files(path.join(workspace, folder)).map(relative => [path.join(folder, relative), digest(path.join(workspace, folder, relative))])));
const allowedEnv = /^(path|pathext|systemroot|systemdrive|windir|comspec|temp|tmp|userprofile|homedrive|homepath|home|appdata|localappdata|programfiles(?:\(x86\))?|commonprogramfiles(?:\(x86\))?|lang|lc_all|AGENT_WORKBENCH_TEST_PYTHON|AGENT_WORKBENCH_DEEPSEEK_ROOT|AGENT_WORKBENCH_PURGE_ROOT)$/i;
const env = Object.fromEntries(Object.entries(process.env).filter(([key]) => allowedEnv.test(key)));
const inheritedPath = Object.entries(env).find(([key]) => key.toLowerCase() === 'path')?.[1] || '';
@@ -27,9 +28,9 @@ for (const key of Object.keys(env)) if (key.toLowerCase() === 'path') delete env
env.PATH = [path.dirname(process.execPath), inheritedPath].filter(Boolean).join(path.delimiter);
const directory = path.join(sourceRoot, '.qa', 'package-validation-' + randomUUID());
fs.mkdirSync(directory, { recursive: true });
const inputFolders = ['src', 'tests', 'scripts'];
const inputFolders = ['src', 'tests', 'scripts', ...['benchmarks', 'public', 'examples', 'research'].filter(folder => fs.existsSync(path.join(sourceRoot, folder)))];
const inputNames = [...inputFolders.flatMap(folder => files(path.join(sourceRoot, folder)).map(relative => path.join(folder, relative))),
'package.json', 'package-lock.json', 'vite.config.ts', 'tsconfig.json', 'index.html'].sort();
'package.json', 'package-lock.json', 'vite.config.ts', 'tsconfig.json', 'index.html', ...['.gitattributes', '.gitignore', 'Start-Desktop.cmd', 'Start-Workbench.cmd', 'Stop-Workbench.cmd'].filter(name => fs.existsSync(path.join(sourceRoot, name)))].sort();
const snapshot = () => inputNames.map(relative => ({ path: relative.replaceAll('\\', '/'), bytes: fs.statSync(path.join(sourceRoot, relative)).size, sha256: digest(path.join(sourceRoot, relative)) }));
const inputs = snapshot(); fs.writeFileSync(path.join(directory, 'inputs.json'), JSON.stringify(inputs, null, 2));
let stage = 'environment preflight', python = null;
@@ -76,13 +77,13 @@ try {
}
console.log(JSON.stringify({ sourceRoot, directory, nodeVersion: process.version, python }));
stage = 'install';
await run([npmCli, 'ci', '--offline', '--no-audit', '--no-fund'], 'install.log');
await run([npmCli, 'ci', '--offline', '--ignore-scripts', '--no-audit', '--no-fund'], 'install.log');
stage = 'build'; await run([npmCli, 'run', 'build'], 'build.log');
const actualFiles = ['dist-client', 'dist-server'].flatMap(folder => files(path.join(sourceRoot, folder)).map(relative => path.join(folder, relative)));
const actualFiles = buildFolders.flatMap(folder => files(path.join(sourceRoot, folder)).map(relative => path.join(folder, relative)));
assert.deepEqual(actualFiles.slice().sort(), Object.keys(expected).sort());
for (const [relative, sha256] of Object.entries(expected)) assert.equal(digest(path.join(sourceRoot, relative)), sha256, relative);
let actualRea = null;
if (['1.8.0', '1.9.0', '1.10.0', '1.10.1', '1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0'].includes(evidence.version)) {
if (['1.8.0', '1.9.0', '1.10.0', '1.10.1', '1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0', '1.15.0'].includes(evidence.version)) {
stage = 'contained REA';
const reaOutput = await run(['scripts/verify-rea.mjs', '--app-root', sourceRoot, '--contained'], 'rea.log');
actualRea = JSON.parse(reaOutput.trim().split('\n').at(-1));
@@ -90,7 +91,7 @@ try {
assert.ok(actualRea.results.every(item => item.rootExited === true));
}
let actualEngine = null;
if (['1.9.0', '1.10.0', '1.10.1', '1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0'].includes(evidence.version) && process.argv.includes('--native-engines')) {
if (['1.9.0', '1.10.0', '1.10.1', '1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0', '1.15.0'].includes(evidence.version) && process.argv.includes('--native-engines')) {
stage = 'native DeepSeek SDK';
const output = await run(['scripts/verify-engines.mjs'], 'engines.log');
actualEngine = JSON.parse(output.trim().split('\n').at(-1));
@@ -98,7 +99,7 @@ try {
assert.equal(actualEngine.modelRequests, 0); assert.equal(actualEngine.rootExited, true); assert.equal(actualEngine.nativeMethods.length, 6);
}
let actualBrowser = null;
if (['1.10.0', '1.10.1', '1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0'].includes(evidence.version)) {
if (['1.10.0', '1.10.1', '1.11.0', '1.12.0', '1.12.1', '1.13.0', '1.14.0', '1.15.0'].includes(evidence.version)) {
stage = 'contained browser runtime';
const output = await run(['scripts/verify-browser.mjs', '--app-root', sourceRoot, '--contained'], 'browser.log');
actualBrowser = JSON.parse(output.trim().split('\n').at(-1));
@@ -108,7 +109,7 @@ try {
assert.equal(actualBrowser.modelRequests, 0);
}
let actualControlled = null;
if (evidence.version === '1.14.0') {
if (['1.14.0', '1.15.0'].includes(evidence.version)) {
stage = 'compiled controlled-engine delivery';
const output = await run(['scripts/verify-controlled-release.mjs', '--app-root', sourceRoot], 'controlled-engine.log');
actualControlled = JSON.parse(output.trim().split('\n').at(-1));
@@ -119,9 +120,23 @@ try {
const output = await run(['--test', '--test-concurrency=2', '--test-reporter=tap', ...files(path.join(sourceRoot, 'tests')).filter(file => !file.includes(path.sep) && file.endsWith('.test.mjs')).map(file => path.join('tests', file))], 'tests.tap');
const stats = Object.fromEntries([...output.matchAll(/^# (tests|pass|fail|cancelled|skipped|todo) (\d+)\s*$/gm)].map(match => [match[1], Number(match[2])]));
assert.ok(stats.tests > 0); assert.equal(stats.pass, stats.tests); for (const key of ['fail', 'cancelled', 'skipped', 'todo']) assert.equal(stats[key], 0);
let actualDesktop = null;
if (evidence.version === '1.15.0') {
stage = 'actual contained Electron window'; await run(['scripts/verify-desktop-window.mjs'], 'desktop-window.log');
const window = JSON.parse(fs.readFileSync(path.join(sourceRoot, '.qa/desktop-window-latest.json'), 'utf8'));
assert.equal(window.success, true); assert.equal(path.resolve(window.appRoot), sourceRoot); assert.equal(window.results.length, 7);
stage = 'actual contained Electron engine UI'; await run(['scripts/verify-controlled-ui.mjs', '--desktop'], 'desktop-controlled-ui.log');
const uiLocation = JSON.parse(fs.readFileSync(path.join(sourceRoot, '.qa/desktop-controlled-ui-location.json'), 'utf8'));
const ui = JSON.parse(fs.readFileSync(uiLocation.result, 'utf8')); assert.equal(ui.passed, true); assert.equal(ui.testedVersion, evidence.version); assert.equal(ui.assertions.length, 9);
stage = 'actual contained Electron engineering projects'; await run(['scripts/verify-controlled-projects.mjs', '--desktop'], 'desktop-projects.log');
const projectLocation = JSON.parse(fs.readFileSync(path.join(sourceRoot, '.qa/desktop-projects-location.json'), 'utf8'));
const projects = JSON.parse(fs.readFileSync(projectLocation.result, 'utf8')); assert.equal(projects.success, true); assert.equal(projects.testedVersion, evidence.version); assert.equal(projects.cases.length, 16);
assert.deepEqual(projects.results.map(value => value.parentCheck.results[0].tests.passed), [8, 9, 10]);
actualDesktop = { window, ui, projects, scope: 'Independently rebuilt source and its shipped Electron binary; original window/engine/engineering assertions. Own credentials/HTTP fixtures; Windows native picker and complete host crash/stop have separate actual evidence.' };
}
assert.deepEqual(snapshot(), inputs, 'Source package inputs changed during validation.');
const result = { success: true, version: evidence.version, testedAt: new Date().toISOString(), sourceRoot, directory, nodeVersion: process.version, python, offlineInstall: true,
inputs, inputFiles: inputs.length, logs, buildMatchedFiles: Object.keys(expected).length, tests: stats, actualRea, actualEngine, actualBrowser, actualControlled, credentialsHandling: 'Only build/runtime environment paths are inherited; no configured model credentials are used. The controlled-engine stage supplies a declared owned local response fixture.' };
inputs, inputFiles: inputs.length, logs, buildMatchedFiles: Object.keys(expected).length, buildFiles: Object.entries(expected).map(([path, sha256]) => ({ path: path.replaceAll('\\', '/'), sha256 })), tests: stats, actualRea, actualEngine, actualBrowser, actualControlled, actualDesktop, credentialsHandling: 'Only build/runtime environment paths are inherited; no configured model credentials are used. The controlled-engine stage supplies a declared owned local response fixture.' };
save(result); console.log(JSON.stringify({ success: true, version: result.version, sourceRoot, directory, tests: stats, nodeVersion: process.version, python, inputFiles: inputs.length, buildMatchedFiles: result.buildMatchedFiles, rea: !!actualRea?.success, engine: !!actualEngine?.success, browser: !!actualBrowser?.success, controlled: !!actualControlled?.success }));
} catch (error) {
save({ success: false, version: evidence.version, testedAt: new Date().toISOString(), stage, error: error.message, sourceRoot, directory, nodeVersion: process.version, python, inputs, logs });
+1 -1
View File
@@ -8,7 +8,7 @@ import { workerHttp } from '../tests/fixtures/workers-http.mjs';
import { workerDefinition, ownerBytes, applyWorkerProposals, approveWorkerChecks } from '../tests/fixtures/workers-project.mjs';
if (process.argv.includes('--help')) { console.log('node scripts/verify-workers-ui.mjs — owned local fixture, real isolated Git/checks/review, desktop/390px, conflicts, stop, recovery and late source errors'); process.exit(0); }
const version = JSON.parse(fs.readFileSync('package.json', 'utf8')).version, prefix = version === '1.14.0' ? 'controlled-engine-workers' : version === '1.13.0' ? 'workers' : 'workers-development';
const version = JSON.parse(fs.readFileSync('package.json', 'utf8')).version, prefix = version === '1.15.0' ? 'desktop-host-workers' : version === '1.14.0' ? 'controlled-engine-workers' : version === '1.13.0' ? 'workers' : 'workers-development';
const directory = path.resolve('.qa', 'workers-ui-' + randomUUID()); fs.mkdirSync(directory, { recursive: true });
const disposers = [], t = { after: dispose => disposers.push(dispose) }, f = fixture(t), host = await workerHttp(f);
assert.notEqual(new URL(host.url).port, '4318');
+1 -1
View File
@@ -25,7 +25,7 @@ export function inspectHostInstallation(directory, { full = false } = {}) {
names.set(row.path, row);
if (full || /^(dist-client|dist-server|dist-desktop)\//.test(row.path)) files.add(row.path);
}
if (!manifest) for (const entry of fs.readdirSync(path.join(root, 'dist-client/assets'))) files.add('dist-client/assets/' + entry);
if (!manifest || manifest.kind === 'source') for (const entry of fs.readdirSync(path.join(root, 'dist-client/assets'))) files.add('dist-client/assets/' + entry);
const inputs = [...files].sort().map(relative => {
const file = ownedChild(root, relative), identity = binaryIdentity(file), declared = names.get(relative);
if (manifest?.kind === 'portable' && !declared) throw hostError('实际宿主程序不在原发行清单中。');
+1 -1
View File
@@ -1,6 +1,6 @@
import { randomUUID, createHash } from 'node:crypto';
export const APP_VERSION = '1.14.0';
export const APP_VERSION = '1.15.0';
export function executionEnvironment() {
const windows = process.platform === 'win32';
+3 -3
View File
@@ -7,9 +7,9 @@
1. [complete] A:Windows实际SID/创建/程序/安装/数据目录与签名挑战,注册/并发锁/启动意图;49/49相关专项和9组真实隔离主机通过。原中文ACL编码/重复权限失败及验收字段失败保留,修正后新证据独立保存。关闭真实窗口留待C/E,A不宣称桌面交付;正式PID23172/1.14保持。
2. [complete] B:CurrentUser DPAPI密文与registry同一SQLite事务,显式新填/确认导入、保存/关闭/删除/清除及失败回退;21/21专项兼容、7组实际浏览器/重启/HTTP/明文扫描通过。实际PID28724→49924、原配置/密文/6表原行保持,2条恢复Key实际3次夹具响应通过。不同Windows账号未实际切换,仅SID拒绝/错误entropy/损坏实际验证;原UI无标签/环境依赖失败及布局截图保留。
3. [complete] C:Electron45.2实际隔离窗口、6固定preload与所属本机入口;17相关专项、7组真实自动流程和独立原生Windows目录选择框通过。原Task/Run/审批/eventSeq/完成工具保持,1440/390视觉核对;所有参数/PID/导航/状态/动作等待和遮罩原失败保留。
4. [in_progress] D:固定依赖/清单、Windows安装与独立新目录更新、活动阻挡和失败回退;实际安装/更新/卸载演练及持久数据核对。安装管理器草稿未构建/验收,不计通过。
5. [pending] E:实际桌面Node/React/Python工程、原失败修复/审批/父审查/检查/Git与窗口/宿主生命周期及unknown不重放。
6. [pending] F:冻结代码/基准、完整回归/全部界面、独立源码/桌面安装/便携与保留状态升级,封存原日志/截图/失败、回顾和下一阶段计划。
4. [complete] D:固定依赖/清单、Windows安装、根启动器与版本指针更新/故障回退、活动阻挡和全数据巡检;16项最终检查、6组巡检UI、11组实际安装演练通过,111份原证据保持,提交f7a5969。
5. [complete] E:实际桌面Node/React/Python三个工程16节点、父8/8、9/9、10/10及构建/编译/审查/Git,原生生命周期UI9组通过;339输入和67份原失败/成功证据保持,提交b77783b。
6. [in_progress] F:1.15版本与独立证据路径、离线源码Electron运行和全部构建范围接入;构建及17相关检查通过。接着冻结代码/基准、完整历史UI与独立源码、最终桌面安装/便携和保留状态正式升级,封存原失败/回顾及PLAN-1.16。
初读确认:现有server与窗口无独立桌面入口,Start-Workbench只凭健康响应复用端口;正式1.14宿主路径和当前runtimeRoot不同。凭据仅内存,ModelHub/ConfigImporter已有事务回退基础。npm核对Electron稳定45.2.0要求Node>=22.12,现固定Node24.22.1满足;安装/运行二进制仍待独立验证。首次大批读取被限额截断,后续只按真实函数及小段读取,不把未返回源码视为已核对。
+16 -1
View File
@@ -2,12 +2,27 @@ import test from 'node:test';
import assert from 'node:assert/strict';
import fs from 'node:fs';
import path from 'node:path';
import { randomBytes } from 'node:crypto';
import { randomBytes, createHash } from 'node:crypto';
import { fixture, workspace, until } from './helpers.mjs';
import { hostProof, verifyHostProof, assertHostStop } from '../src/shared/contracts/desktop-host.mjs';
import { ownedChild, writeState, readState } from '../src/core/desktop/file-state.mjs';
import { HostAuthority } from '../src/core/desktop/host-authority.mjs';
import { Provider } from '../src/server/provider.mjs';
import { inspectHostInstallation } from '../src/core/desktop/app-installation.mjs';
test('a rebuilt source archive binds actual generated client assets and retains its original running identity after disk changes', t => {
const f = fixture(t), root = path.join(f.root, 'owned-source-build'); fs.mkdirSync(root);
for (const directory of ['dist-client', 'dist-server', 'dist-desktop']) fs.cpSync(path.join(workspace, directory), path.join(root, directory), { recursive: true });
const packageBytes = fs.readFileSync(path.join(workspace, 'package.json')); fs.writeFileSync(path.join(root, 'package.json'), packageBytes);
fs.writeFileSync(path.join(root, 'MANIFEST.json'), JSON.stringify({ app: 'agent-workbench', kind: 'source', version: JSON.parse(packageBytes).version,
files: [{ path: 'package.json', bytes: packageBytes.length, sha256: createHash('sha256').update(packageBytes).digest('hex') }] }));
const authority = new HostAuthority({ appDirectory: root, store: f.store, provider: new Provider(), runtime: { appDirectory: root, active: new Map(), waiters: new Map() }, server: { address: () => ({ port: 49998 }) } });
const running = authority.identity(), asset = 'dist-client/assets/' + fs.readdirSync(path.join(root, 'dist-client/assets')).find(name => name.endsWith('.js'));
assert.ok(inspectHostInstallation(root).inputs.some(value => value.path === asset));
fs.appendFileSync(path.join(root, asset), '\n// owned source-archive disk change\n');
assert.equal(authority.identity().runtimeIdentity, running.runtimeIdentity);
assert.notEqual(inspectHostInstallation(root).identitySha256, running.runtimeIdentity);
});
test('host challenge proof rejects a stale challenge, changed identity and changed signature', () => {
const identity = { instanceId: 'owned', pid: 123, createdAt: 'fixed', runtimeIdentity: 'fixed-input' }, secret = randomBytes(32), challenge = randomBytes(32).toString('hex');