docs: record 1.3 precision delivery and release evidence

This commit is contained in:
Agent Workbench Dev
2026-10-29 18:16:42 +08:00
parent 14ce622028
commit f1fe9d39fc
15 changed files with 875 additions and 7 deletions
+1 -1
View File
@@ -16,4 +16,4 @@
桌面及390px的索引、文件行与精确差异截图已目视核对,见[screenshots](screenshots/1.3-index-desktop.png)。初始两项Windows Node HTTP连接重置已独立复验并采用有限QA请求独立连接,完整112项最终回归全部通过;没有自动重试副作用请求。
便携包、独立源码安装/构建、原进程升级与真实模型长文件精确交付的最终结果汇总到[VALIDATION](VALIDATION.md)。真实模型样本增加420行受保护说明及两种BOM/换行格式,要求用索引定位正文、仅读取必要源码并提出精确补丁,沿原Task完成14项测试、Git和报告;不由单个样本推断完整M2/M6成功率。
便携包启动/复用/随附Node与npm/优雅停止、源码ZIP独立离线安装/构建/112项回归,以及原PID23080保留模型连接的升级均已通过,结果见[VALIDATION](VALIDATION.md)。真实模型长文件样本119秒通过:420行说明和BOM/换行格式逐字节保留,索引定位正文、必要行读取、四份精确补丁,沿原Task完成14项测试、build、工作台Git和报告。用量15次请求/280253个已报告tokens;尚无同题成本对照,不宣称索引已节省用量。完整M2/M6验收继续实施。
+22
View File
@@ -1,5 +1,27 @@
# Agent Workbench 验收记录
## 1.3.0 · 工程索引与精确补丁
Windows x64、Node24.22.1、Edge156.0.4314.42。构建和112/112项行为/API/实际进程退出与连续升级检查通过,0失败/跳过/取消。基础10、诊断9、任务7、Git交付9和索引8共43组真实界面流程全部通过,页面/请求异常为零;桌面与390px截图已目视核对。
| 验证 | 结果 | 结构化证据 |
| --- | --- | --- |
| AST/候选索引、增量依赖、精确补丁及原有恢复契约 | 112/112通过 | [m2-index-tests-validation.json](m2-index-tests-validation.json) |
| 基础/诊断/任务/Git/索引界面 | 10/9/7/9/8组全部通过 | [QA-1.3.md](QA-1.3.md) |
| 便携包与随附Node/npm | 独立启动/复用/停止、4项实际示例检查通过 | [m2-index-release-validation.json](m2-index-release-validation.json) |
| 源码ZIP独立安装/构建/回归 | 离线安装、6份构建文件一致、112/112通过 | [m2-index-source-package-validation.json](m2-index-source-package-validation.json) |
| 两个ZIP的CRC/路径/清单/散列 | 全部通过 | [m2-index-archives-validation.json](m2-index-archives-validation.json) |
| 正式原进程升级 | 1.3.0、PID23080/内存连接及历史状态保留 | [m2-index-upgrade-live-validation.json](m2-index-upgrade-live-validation.json) |
| 真实模型精确补丁交付 | 13失败→14通过、build、四文件commit和报告 | [m2-index-live-model-validation.json](m2-index-live-model-validation.json) |
实际配置的gpt-6.1-sol用119秒、1Task/2Run、1次四文件审查完成长文件样本。模型调用1次index_project、2次query_code,仅读取各文件421行后的必要源码,四文件均用propose_patch。每个源文件420行受保护说明、normalize/cli的BOM与CRLF、summary/render的LF、10个保护输入、其他工作区与暂存修改均保留。封存14项测试由1通过/13失败变为14通过,build和构建CLI结果正确。
模型准备Git预览;验收器审核提案、批准封存的npm脚本和确认工作台预览,工作台自身生成并核对样本commit `5dec27e38ad9b3a2c5da53debb8af29884cec690`及[Markdown报告](m2-index-model-delivery-report.md)/[JSON报告](m2-index-model-delivery-report.json)。[实际diff](m2-index-live-model.diff)保存四文件的局部修改。验收器未编写修复实现或自行执行Git提交,未读取、复制或导出模型Key。
实际用量为15次请求,输入277387、输出2866、合计280253个已报告tokens,未知用量请求0;未配置单价不换算费用。这个样本证明定位和格式保留的正确性;尚未建立上下文成本下降的同题对照,不能据此宣称节省tokens。上下文投影与预算将继续优化。
这是一个合成Node工程样本;完整M2三类工程、ChangeSet/隔离worktree、外部引擎、Electron及手机配对继续实施。JS/TS/React索引是静态声明/导入证据,Python为声明候选,均不表示完整编译器引用或运行时依赖解析。首轮两项Windows HTTP连接重置已独立复验;有限QA请求使用独立连接,不重试副作用请求,最终完整回归通过。
## 1.2.0 · 当前Task的Git交付与证据报告
Windows x64、Node24.22.1、Edge156.0.4314.42。构建通过;完整92项行为/API/实际退出恢复测试全部通过,0失败/跳过/取消。首轮90/92的两个失败已定位并修复,15项相关专项及最终完整回归均通过。
+4 -4
View File
@@ -1,14 +1,14 @@
{
"ready": true,
"url": "http://127.0.0.1:4318",
"version": "1.2.0",
"version": "1.3.0",
"mode": "compatible",
"model": "gpt-6.1-sol",
"hasKey": true,
"pid": 23080,
"appRoot": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\releases\\Agent-Workbench-1.2.0-win-x64-20261029085728391",
"appRoot": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\releases\\Agent-Workbench-1.3.0-win-x64",
"originalAppRoot": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\releases\\Agent-Workbench-0.1.0-win-x64",
"dataRoot": "C:\\Users\\wangy\\AppData\\Local\\AgentWorkbench",
"upgradedAt": "2026-10-29T08:58:58.364Z",
"upgradedAt": "2026-10-29T10:05:56.806Z",
"storage": "sqlite"
}
}
+24
View File
@@ -0,0 +1,24 @@
{
"success": true,
"version": "1.3.0",
"testedAt": "2026-10-29T10:00:17.501393+00:00",
"archives": [
{
"kind": "portable",
"path": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\releases\\Agent-Workbench-1.3.0-win-x64.zip",
"files": 2130,
"crcPassed": true,
"manifestAndDirectoryMatch": true,
"sha256": "6feab9f89ac8b0658836c4ca1983b482693e30eadfa2620c30f0a2e5f545d153"
},
{
"kind": "source",
"path": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\releases\\Agent-Workbench-1.3.0-source.zip",
"files": 275,
"crcPassed": true,
"manifestAndDirectoryMatch": true,
"sha256": "4cbbad34e0e80ca437020baa3ca331ab67666b595e185d7a3a5bcfeeabf4b620"
}
],
"extractionRoot": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\source-check-9e1b2f02ba574915ae640186d87f0e3c\\Agent-Workbench-1.3.0-source"
}
File diff suppressed because one or more lines are too long
+101
View File
@@ -0,0 +1,101 @@
diff --git a/src/cli.mjs b/src/cli.mjs
index 42aea37..84ccace 100644
--- a/src/cli.mjs
+++ b/src/cli.mjs
@@ -423,6 +423,16 @@ import { normalizeTasks } from './normalize.mjs';
import { summarize } from './summary.mjs';
import { renderText } from './render.mjs';
try {
- const tasks = normalizeTasks(JSON.parse(fs.readFileSync(process.argv[2], 'utf8')));
- console.log(renderText(summarize(tasks)));
-} catch (error) { console.error(error.message); }
+ const args = process.argv.slice(2);
+ if (args.length < 1 || args.length > 2 || !args[0] || args[0].startsWith('-') ||
+ (args.length === 2 && args[1] !== '--json')) {
+ throw new TypeError('用法:node src/cli.mjs <JSON文件> [--json]');
+ }
+ const tasks = normalizeTasks(JSON.parse(fs.readFileSync(args[0], 'utf8')));
+ const summary = summarize(tasks);
+ const output = args[1] === '--json' ? JSON.stringify(summary) : renderText(summary);
+ process.stdout.write(output + '\n');
+} catch (error) {
+ process.exitCode = 2;
+ process.stderr.write('输入错误:' + error.message + '\n');
+}
diff --git a/src/normalize.mjs b/src/normalize.mjs
index 5a46e5a..84107b9 100644
--- a/src/normalize.mjs
+++ b/src/normalize.mjs
@@ -419,6 +419,29 @@
// 保留工程说明 419 · src/normalize.mjs · 原文🙂
// 保留工程说明 420 · src/normalize.mjs · 原文🙂
export function normalizeTasks(input) {
- if (!Array.isArray(input)) throw new TypeError('tasks must be an array');
- return input;
+ if (!Array.isArray(input)) throw new TypeError('任务必须是数组');
+ if (input.length > 10000) throw new RangeError('任务最多10000项');
+ const ids = new Set();
+ const statuses = new Set(['todo', 'doing', 'done', 'blocked']);
+ const result = [];
+ for (const task of input) {
+ if (task === null || typeof task !== 'object' || Array.isArray(task)) {
+ throw new TypeError('每项任务必须是对象');
+ }
+ if (typeof task.id !== 'string' || typeof task.title !== 'string') {
+ throw new TypeError('id和title必须是非空字符串');
+ }
+ const id = task.id.trim();
+ const title = task.title.trim();
+ if (!id || !title) throw new TypeError('id和title不能为空');
+ if (ids.has(id)) throw new TypeError('任务id不能重复');
+ const { status, minutes } = task;
+ if (!statuses.has(status)) throw new TypeError('任务状态无效');
+ if (!Number.isSafeInteger(minutes) || minutes < 0 || minutes > 1000000) {
+ throw new RangeError('minutes必须是0到1000000之间的安全整数');
+ }
+ ids.add(id);
+ result.push({ id, title, status, minutes });
+ }
+ return result;
}
diff --git a/src/render.mjs b/src/render.mjs
index 5cacc86..2752ba6 100644
--- a/src/render.mjs
+++ b/src/render.mjs
@@ -419,5 +419,10 @@
// 保留工程说明 419 · src/render.mjs · 原文🙂
// 保留工程说明 420 · src/render.mjs · 原文🙂
export function renderText(summary) {
- return `Done: ${summary.counts.done}/${summary.total}`;
+ const { total, counts, completedPercent, remainingMinutes } = summary;
+ return [
+ `交付进度:${counts.done}/${total}(${completedPercent}%)`,
+ `待办 ${counts.todo} · 进行中 ${counts.doing} · 已完成 ${counts.done} · 阻塞 ${counts.blocked}`,
+ `剩余预计:${remainingMinutes} 分钟`
+ ].join('\n');
}
diff --git a/src/summary.mjs b/src/summary.mjs
index 6679d91..29d07be 100644
--- a/src/summary.mjs
+++ b/src/summary.mjs
@@ -419,7 +419,17 @@
// 保留工程说明 419 · src/summary.mjs · 原文🙂
// 保留工程说明 420 · src/summary.mjs · 原文🙂
export function summarize(tasks) {
- const done = tasks.filter(task => task.status === 'done').length;
- return { total: tasks.length, counts: { done }, completedPercent: tasks.length ? 100 : 0,
- remainingMinutes: tasks.reduce((sum, task) => sum + task.minutes, 0) };
+ const counts = { todo: 0, doing: 0, done: 0, blocked: 0 };
+ let remainingMinutes = 0;
+ for (const task of tasks) {
+ counts[task.status] += 1;
+ if (task.status !== 'done') remainingMinutes += task.minutes;
+ }
+ const total = tasks.length;
+ return {
+ total,
+ counts,
+ completedPercent: total ? Math.round(counts.done / total * 100) : 0,
+ remainingMinutes
+ };
}
File diff suppressed because one or more lines are too long
+57
View File
@@ -0,0 +1,57 @@
# 项目交付报告
**检查通过并已提交**
生成时间:2026-10-29T10:08:50.011Z
项目:精确补丁交付工程
项目目录:D:\Ai-project\REA逆向\Agent-Workbench\.qa\live-precision-62378407-9a63-4df8-b1a4-c3968684c9d4\精确补丁交付工程
Task:2cc4aa20-2798-44ea-a5ea-fae141d19d0f
## 任务与范围
完成当前 Delivery Report 工程,严格实现 README.md 的七项要求。先用 inspect_project 识别工程,再用 plan_checks 选择 test/build,并用 run_checks 实际运行封存输入,取得失败的测试和日志。依据实际失败读取有关文件,使用 read_artifact 进一步核对失败日志。先用 index_project 建立工程索引,再用 query_code 的 symbols/imports 结果定位源码,特别是CLI入口的实际导入行。每个源码模块前420行都是须原字节保留的说明;用 read_file 指定start_line至少421的必要行范围,取得最新SHA256。已有源码只能用 propose_patch 及该SHA256、唯一old_text锚点精确修改;保留BOM及原LF/CRLF,不使用 propose_file,不读整个说明块。允许修改的源码仅为 src/normalize.mjs、src/summary.mjs、src/render.mjs、src/cli.mjs。测试、fixtures、构建脚本、package.json、README 和 OWNER 文件必须保留。四模块都需修复,不能修改验收来通过。本工程没有第三方依赖,不用安装。不要使用 run_command、不要自己提交Git;本轮终端仅通过已识别的npm检查计划运行。提案等待审查应用,随后继续同一目标并实际重新运行 test/build,依据日志修复直至通过。检查通过后用 plan_delivery 准备本Task的四文件本地Git预览,等待用户独立确认;最后用中文报告实际测试数量、构建结果和改动。
源码指纹:c37b327ad73343def8f30cd01d0edef7de8f62f2bcb35107aaa9d693ad1aff81
当前Git HEAD:5dec27e38ad9b3a2c5da53debb8af29884cec690
分支:main
源码和检查证据不代替业务需求验收。
## 改动
| 文件 | 状态 | SHA256 |
| --- | --- | --- |
| src/normalize.mjs | applied | f04ef9fdad8b2eb7ca02d38e4e587885a42ee731b6971002357b9659e7fbf8d8 |
| src/summary.mjs | applied | 06c0bb1d5ae472c343e8137d05c175c7216ca7a5c8c1c6c0c7dcba0df38fbb13 |
| src/render.mjs | applied | 2882b485d686e1cbdaf8de2834592f9a84921ed63340dbae42916145df907ca7 |
| src/cli.mjs | applied | 038a0ea3e1ac49ca0f9474ed3238f6cc65e4d667e1d4b6fa64d9d2e54ff2d0c1 |
## 实际检查
| Job / 命令 | 状态 | 退出码 | 实际测试 | 日志SHA256 |
| --- | --- | --- | --- | --- |
| 01cb9efa / npm run test | 源码已变化 / failed | 1 | 1/14 | fd9fda5ef6e35159f1f33e908bdfb504b759d7153d0e0aeeb33c1fcef28ced10 |
| 01cb9efa / npm run build | 源码已变化 / passed | 0 | 不适用或未知 | 38566d6bd1635a690931ce24b85aa15a188dc2844ff8d886289985039a9f9be5 |
| 1ff8446c / npm run test | passed | 0 | 14/14 | 79228e3ae61f187ec198bbdf6f19c95fe2058b3a317ed2aeea815883db8db3de |
| 1ff8446c / npm run build | passed | 0 | 不适用或未知 | 38566d6bd1635a690931ce24b85aa15a188dc2844ff8d886289985039a9f9be5 |
## Git交付
- completed · 5dec27e38ad9b3a2c5da53debb8af29884cec690 · src/cli.mjs, src/normalize.mjs, src/render.mjs, src/summary.mjs
已生成并核对本地提交;任务文件、实际检查和其他工作区/暂存改动均已保留。
## 模型与用量
| Run | 模型 | 状态 | 已报告tokens |
| --- | --- | --- | --- |
| 027f0396-b8fc-4bff-9b73-ddae3efd932f | gpt-6.1-sol | needs_review | 124986 |
| f33d3132-1954-4a34-95ff-1145660cc84b | gpt-6.1-sol | completed | 155267 |
累计已报告:280253 tokens;未报告请求:0。费用:未知。
## 尚待确认
- 本切片验证已选工程检查和提交范围;业务需求验收另行记录。
- 提交为本地普通提交,不执行仓库hooks和GPG签名。
- 未配置模型单价,费用未知。
+34
View File
@@ -0,0 +1,34 @@
{
"success": true,
"date": "2026-10-29T10:00:38.764Z",
"appRoot": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\releases\\Agent-Workbench-1.3.0-win-x64",
"version": "1.3.0",
"nodeVersion": "v24.22.1",
"verifiedFiles": 2130,
"url": "http://127.0.0.1:51900",
"reusedPid": 47112,
"checks": {
"jobId": "d1578f3f-e1ca-4915-9f8a-32cee13e6a5f",
"status": "passed",
"tests": {
"total": 4,
"passed": 4,
"failed": 0,
"cancelled": 0,
"skipped": 0
}
},
"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",
"report availability",
"graceful stop"
],
"realModel": false
}
@@ -0,0 +1,17 @@
{
"success": true,
"version": "1.3.0",
"testedAt": "2026-10-29T10:03:52.533Z",
"sourceRoot": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\.qa\\source-check-9e1b2f02ba574915ae640186d87f0e3c\\Agent-Workbench-1.3.0-source",
"offlineInstall": true,
"buildMatchedFiles": 6,
"tests": {
"tests": 112,
"pass": 112,
"fail": 0,
"cancelled": 0,
"skipped": 0,
"todo": 0
},
"credentialsHandling": "Only build/runtime environment paths are inherited; no model configuration or credentials are used."
}
@@ -0,0 +1,83 @@
{
"success": true,
"version": "1.3.0",
"records": [
{
"version": "1.3.0",
"pid": 23080,
"hasKey": true,
"projects": 5,
"sessions": 13,
"appDirectory": "D:\\Ai-project\\REA逆向\\Agent-Workbench\\releases\\Agent-Workbench-1.3.0-win-x64",
"verified": true,
"connectionPreserved": true,
"url": "http://127.0.0.1:4318",
"testedAt": "2026-10-29T10:05:56.806Z",
"previousVersion": "1.2.0",
"model": "gpt-6.1-sol",
"preservedSessions": [
{
"id": "c32c9c87-8239-4606-a4c0-783529603bc7",
"projectId": "82850b57-b496-4f1c-912c-9c54d1858917",
"status": "completed",
"revision": 209,
"messages": "2d578dee66c9da55e5ac675f1ee30fd6ee9955d835bca4039934c0073d9df4c7",
"events": "41a8dff99d153860971a47750fc48c181b06fd7d6d476b3123055d36658094c1",
"changes": "910bf7865f3b3f357edbc7a2f42e851eb4c3ae1c4923bf8e1d9e3f7e8db72205",
"execution": "35647b0b3eea1b171a765bbc20badbdd7f4db1495436a7d788867031d83cda86"
},
{
"id": "1cf22a27-ac13-4c95-a878-9b4356ccb5b3",
"projectId": "3423bff0-d418-4a6d-9cc7-66bd85875a51",
"status": "completed",
"revision": 34,
"messages": "b91dc3a60b34051b3df0bc501372e15a614eafe3594943b8da88addd0557af73",
"events": "41e7056c3f824a03d9d1b5abd4ee3868d964ecb144e33642bc9bbd67df987bfb",
"changes": "4f53cda18c2baa0c0354bb5f9a3ecbe5ed12ab4d8e11ba873c2f11161202b945",
"execution": "f6ba74f339fae6a5d301ddf6440fd6494309364308430c758ca36de21d8e8a6d"
},
{
"id": "e8846c6e-1ba2-497e-b0db-ca5778181eb0",
"projectId": "3423bff0-d418-4a6d-9cc7-66bd85875a51",
"status": "completed",
"revision": 182,
"messages": "396c60e8f21d630f5a19867a24f8e9fe36208ac24c55e65a581d499e3542fb91",
"events": "3fa8b3f0a07275989f80dfc151db767b2e62943ce46f86940a1623585c2c794f",
"changes": "a78838fdd7e236110212b0eea2eb87a59c5d3a00ac6a49a98e2b44f87c1e8069",
"execution": "8c27fe79d5367826f8cc9fd8af582b4140b55eb685f0e7150042d06b8b82e087"
},
{
"id": "72d3d22f-14d9-4bca-bb6a-ab686bc96b26",
"projectId": "ffb29b87-9b26-4bcc-9337-2ba535301c27",
"status": "failed",
"revision": 209,
"messages": "179f776107a25f32670fcf2a6cadc168649892d19ce602d034364af23f24230e",
"events": "7db9260050d3550f4ef333fe8047939e8eccd1e0e1ee648059408a92b04fb293",
"changes": "4f53cda18c2baa0c0354bb5f9a3ecbe5ed12ab4d8e11ba873c2f11161202b945",
"execution": "60edd1935ae62945a4dde1908229b00a37376fa0ba6960a55bbeb8b6f5282980"
},
{
"id": "4abc4201-549e-46e0-947c-3d3f5b53ffae",
"projectId": "b4de81b1-fe73-48d8-a1c4-7f2aee02cf18",
"status": "needs_review",
"revision": 59,
"messages": "2463725fc4dce801dfd6752be582ff53489bc42950a3dd7a6cb917735f90069c",
"events": "594b16c254b0f09cf3f53dab113827a40a08156f263c57365d20b478d26001fa",
"changes": "8fdcbd54c3cafda6bb0a35a19ef244265b6ce4c250017b7d509a69d70c482dde",
"execution": "b8239c1241e17cf78298e20e62a41a13ee609b9a20307e04cdc0ee1da2112323"
},
{
"id": "569c9457-d399-46e6-b53e-4b1fdbb02109",
"projectId": "b4de81b1-fe73-48d8-a1c4-7f2aee02cf18",
"status": "completed",
"revision": 199,
"messages": "6dff812dc2055b1cffa6fbbdc22012bd1de1d716157645fb9df652808584e2ad",
"events": "76fcc4e99e0284c5c147170f17d9dffe88cb6f03f92d44a8a683da9f14beb8b9",
"changes": "e2f726458686db2092bc532514c9ffdbb5cad5c5c586980aa501c60cbc199874",
"execution": "63a02ad224b153236d595e4076aa74e8dfeacb048f9696fe8ec98da37916f614"
}
],
"storage": "sqlite"
}
]
}