agentworkshop 0.7.64 → 0.7.66

package/.output/server/package.json

@@ -1,6 +1,6 @@
 {
   "name": "agentworkshop-prod",
-  "version": "0.7.64",
+  "version": "0.7.66",
   "type": "module",
   "private": true,
   "dependencies": {

package/README-zh.md

@@ -28,7 +28,7 @@
 
 **[English](./README.md)** · **[在线文档](https://kingdol666.github.io/AgentWorkShop)** · **[版本发布](https://github.com/kingdol666/AgentWorkShop/releases)** · **[更新日志](./changelog.md)** · **[插件 API](./docs/plugins.md)** · **[SDK](./docs/sdk.md)**
 
-<sub><b>v0.7.63</b> · 14 个执行引擎 · 6 种现场协议（5 内置 + 串口插件）+ MES REST 集成 · 数据分析台账（产线级权限隔离）· 119 个运行时设置项 · 双语文档（简体中文 / English）</sub>
+<sub><b>v0.7.66</b> · 14 个执行引擎 · 6 种现场协议（5 内置 + 串口插件）+ MES REST 集成（镜像与直连双路径）· 数据分析台账（产线级权限隔离）· 工具调用参数/绑定预检 · 82 个宿主工具 · 119 个运行时设置项 · 双语文档（简体中文 / English）</sub>
 
 <br />
 
@@ -154,7 +154,7 @@ AgentWorkShop 起家于**多智能体软件工作坊**——Channel 内的编码
 | **Channel 群聊与原生 HITL** | 频道时间线是真正的群聊：成员请求升级为**可追踪作业**（不是一条会丢的消息）；带原生 ask 能力的引擎（如 `omp`）把 HITL 提问路由进平台——您批准一次，引擎带着回执继续跑。成员权限与用户通知是一等公民。 |
 | **定时任务** | 任意 Channel 任务可挂上**调度**——固定 `interval`（60s 下限）或 `daily` 每日定点——在 `/workshop/schedules` 页或 REST API 管理，带逐次运行历史、忙等守卫（频道仍有在飞任务时触发顺延）与 Channel 级可见性。 |
 | **持久记忆与频道过程记忆** | 私有 + Channel 共享双域；FTS5 CJK 切分，可选向量混合检索，token 预算注入。其上是**过程记忆**：确定性的 root/child/lead 事件名契约、canonical root 摘要、持久 outbox（死信 + 补偿 worker），让一个目标的来龙去脉跨重启存活。 |
-| **四个入口** | 一个 manager 坐在每扇门后：**WS**（AEP v1 事件流，seq 续传）、**MCP**（约 25 个进程内工具）、**A2A**（JSON-RPC 2.0 + AgentCard）、**REST**。 |
+| **四个入口** | 一个 manager 坐在每扇门后：**WS**（AEP v1 事件流，seq 续传）、**MCP**（38 个进程内工具）、**A2A**（JSON-RPC 2.0 + AgentCard）、**REST**。 |
 
 #### 工业栈
 
@@ -167,9 +167,10 @@ AgentWorkShop 起家于**多智能体软件工作坊**——Channel 内的编码
 | **Recipe 版本化治理** | 参数修改按版本入史（归因 用户/Agent/系统 + 操作者 + 原因）。可非破坏地回退到任意修订版或最近一次良好批次。失效节点参数跳过并明确标识。 |
 | **配方链路优化闭环** | Agent 像人类工艺员一样通过配方管理做优化：`recipe_trial` 把多参数候选**整批试验（不写版本）**→ 复测 → 有进步用 `recipe_update` 固化（同一配方 id，版本+1）→ 劣化触发**统一回退**（`recipe_rollback` 带 `dispatch:true` = 定义回退 + PLC 整批恢复）。单参数直调移出闭环；试验受四层限界与同线节拍卡控。 |
 | **频道绑定产线（只读授权）** | 频道可绑定一条产线：全员无需节点绑定即可读取该线的运维日志、配方历史与实时状态（工具回执标注「频道绑定，只读」），写操作仍需节点授权。产线简报自动注入成员提示词；换绑即时回收成员运行时。 |
-| **数据分析台账（v0.7.62）** | Agent 的分析结果作为一等记录留痕：原生 `analysis_save` / `analysis_query` 宿主工具写入独立的 `analysis_reports` 表，按产线 + 配方 + 时间窗组织，含关键结论、指标、数据来源与操作者归因。台账对 Agent 只增不改；人类经 REST 修订并刷新编辑者归因。专属 `/analysis` 页面经 WS 实时前插新报告，支持产线 / 配方 / 时间 / 关键词多维检索，改删受产线权限门控（`operate` 可改、删除仅 admin）。外部分析插件经同一台账集成——`diag_run` 以后台任务运行 IDD 分析（`stability-report` 产线稳定报告 / `deep-diagnosis` 根因诊断），提交即返 `task_id`，`diag_status` 按taskId 查进度，完成后报告经 `analysis_save` 落账。 |
+| **数据分析台账（v0.7.62）** | Agent 的分析结果作为一等记录留痕：原生 `analysis_save` / `analysis_query` 宿主工具写入独立的 `analysis_reports` 表，按产线 + 配方 + 时间窗组织，含关键结论、指标、数据来源与操作者归因。台账对 Agent 只增不改；人类经 REST 修订并刷新编辑者归因。专属 `/analysis` 页面经 WS 实时前插新报告，支持产线 / 配方 / 时间 / 关键词多维检索，改删受产线权限门控（`operate` 可改、删除仅 admin）。外部分析插件经同一台账集成——`diag_run` 以后台任务运行 IDD **六种分析意图**（深度根因诊断 / 产线稳定报告 / 快速筛查 / DOE 试验分析 / 调参经验挖掘 / 优化战役，按情景注入分析提示词），提交即返 `task_id`，`diag_status` 按 taskId 查进度，完成后报告经 `analysis_save` 落账。 |
 | **工艺参数语义面与调控闭环** | 全部可下发写收敛到**配方面**（`recipe_update` / `recipe_apply` / `recipe_trial`，各受「绑定 → HITL 理由 → 运行门」三门约束）。写入仍被**四层限界**逐层收窄——节点安全量程 ∩ 工艺参数基准限界 ∩ 活动产品限界 ∩ 活动配方工艺窗口——且每次下发自动开一条调控记录，由 `dcw_judge` 落判定（keep / rollback / uncertain），`dcw_rollback` 执行回退。（v0.7.63 摘除了 `param_control` / `param_read` / `dcw_read` 直连节点的遗留死工具面，Agent 工具清单保持干净。） |
 | **产线运营** | 产线 → 产品 → 配方 → 批次。配方窗口门控采集并联锁写入；每条样本打标 `product/recipe/run`，实现按批次隔离。 |
+| **MES REST 集成（双数据路径）** | 生产管理数据进入同一受治理工作空间：`mes-rest` 节点用**声明式请求映射**对接 MES API 的目录/当前值/历史端点（`secretRef`→env token 凭证零明文；钩子可对取数做下沉后处理）。Agent 两种读法——**直连**：`mes_fetch` 按映射发请求实时拿 API 响应（大窗异步分页落 `mesds-*` CSV 数据集，经 `mes_datasets` / `mes_dataset_read` 读取，原始行不进对话）；**镜像**：`http` 数采节点轮询同一 API 入本地 TSDB，`daq_query` 与原生节点同面取数——API 宕机也不断供。`mes_catalog` 按产线严格过滤（fail-closed）。 |
 | **多形态数采帧管线** | 多点轮廓（测厚仪/扫描仪）与 CCD 图像帧流经模板 sink 管线：向量与元数据入 Timescale（`daq_frames`），像素入对象存储（MinIO，不可达时自动降级到本地磁盘）；派生指标阈值越限走既有告警链路。 |
 | **生产化硬化（v0.7.58）** | 保写心跳生命周期根治——失败写不占位（心跳不再直发幻影指令值）、驱动/连接/产线变更即清值、停线自动挂起保写；mock 静默兜底封堵（未知协议显式报错，插件卸载不再假成功）；帧像素 sha256 完整性指纹 + 对象存储按天 GC（删前 DB 交叉校验防误删）；Timescale 按天 chunk + 列存压缩（TSL）；堆水位监控 + 崩溃/失联/水位三重自动重生；写保持窗与试验节拍落盘（重启不绕频控）；grant 运行时复核并入频道绑线（撤权即失活无死角）。 |
 | **Agent 自查工具** | `line_context`、`ops_log`、`recipe_log`、`recipe_versions`、`dcw_journal`——Agent 清楚自己操控的产线/产品/配方，谁做过什么，每个值怎么变。 |
@@ -179,6 +180,7 @@ AgentWorkShop 起家于**多智能体软件工作坊**——Channel 内的编码
 | 能力 | 为何重要 |
 |---|---|
 | **产线级权限** | 工业数据按**产线**三态门控（无权/仅查看/可操控），强制点在数据面——普通用户未授权前看不到任何产线数据。 |
+| **工具调用统一预检（v0.7.66）** | 每次工业工具调用在进入处理器前统一校验参数形状与引用归属：`node_id` / `line_id` / `recipe_id` / `dataset_id` 等必须落在调用者的绑定面（绑定节点 ∪ 频道绑线只读 ∪ 绑定配方所在产线；数据集/作业/模型/计划/场景按谱系回溯）。越界调用直接拒绝，回执写明「你的绑定面 + 下一步」（找谁绑定、改传什么），并把授权性拒绝写入审计日志。 |
 | **全操作审计日志** | 用户 / Agent / 系统 的每个动作都落进同一份可检索日志；操作者归属「Channel名/成员名」，与用户和系统天然区分。经 WS 实时推送。 |
 | **团队级插件开关** | 每个团队（Channel）持有**独立插件开关组**（`channel_plugins`）：被关闭插件的工具不注入该团队 Agent。插件本体经 `aw plugin` 与 `/plugins` 页热管理。 |
 | **插件扩展 API** | `plugins/<name>/` 下的一个自包含目录**同时增强两半**：`index.mjs`（服务端：钩子、路由、Agent 工具、**数采读驱动 / DCW 写驱动**/帧处理器/节点模板、配置分组、KV、定时器）与 `client.mjs`（浏览器：注入具名插槽的面板、i18n、设置 UI）。三种作用域——`builtin`（随包发布）> `project`（检出）> `user`（`~/.AgentWorkShop`）——启停**与代码修改**均有约 1 秒热重载；停用插件的驱动随热重载立即摘除。内置示例 **serial-bridge**（串口通信：读/写驱动 + 串口探针 API + 前端面板）。完整契约见 [`docs/plugins.md`](./docs/plugins.md)。 |
@@ -390,7 +392,7 @@ aw update --check                      # 只报告，不安装
 npm install -g agentworkshop@latest    # 手动等效
 ```
 
-版本遵循 semver。每次 `aw start` 都会校验配置根，并把 `home` 之前的旧版 `data/` 布局迁移进来（以最新文件为准），因此数据可以跨版本存活。SQLite schema 迁移在服务端启动时执行。当前版本：**v0.7.64**——见[版本发布](https://github.com/kingdol666/AgentWorkShop/releases)。
+版本遵循 semver。每次 `aw start` 都会校验配置根，并把 `home` 之前的旧版 `data/` 布局迁移进来（以最新文件为准），因此数据可以跨版本存活。SQLite schema 迁移在服务端启动时执行。当前版本：**v0.7.66**——见[版本发布](https://github.com/kingdol666/AgentWorkShop/releases)。
 
 ### 第一次「Agent × 产线」会话（约 2 分钟）
 
@@ -576,7 +578,7 @@ curl -X POST http://localhost:3000/api/workshop/schedules \
 | 入口 | 端点 | 面向 |
 |---|---|---|
 | **WS** | `/api/workshop/ws?channelId=…` | 仪表盘 / UI——AEP v1 信封，per-channel 单调 `seq`，5000 事件环形缓冲，`lastSeq` 续传，快照兜底。 |
-| **MCP** | 进程内服务，约 25 个工具 | Agent（omp host tools）——管理面 + 作业面工具，Channel 作用域。 |
+| **MCP** | 进程内服务，38 个工具 | Agent（omp host tools）——管理面 + 作业面工具，Channel 作用域。 |
 | **A2A** | `POST /api/workshop/a2a/:agentId/rpc` | 外部 Agent——JSON-RPC 2.0，`AgentCard` 在 `/card`，`tasks/sendSubscribe` SSE。 |
 | **REST** | `/api/workshop/**` | 人 / 脚本——完整管理面。 |
 
@@ -646,7 +648,7 @@ AgentWorkShop/
 │   │   ├── agents/             # AgentInterface：14 个引擎（+ 工业工具）
 │   │   ├── daq/ dcw/ aml/      # 边缘运行时 · 驱动 · 队列 · 存储 · 建模实验室
 │   │   └── db/                 # node:sqlite 仓储层
-│   ├── mcp/                    # MCP 服务（25 个工具，进程内 workshop 面）
+│   ├── mcp/                    # MCP 服务（38 个工具，进程内 workshop 面）
 │   ├── plugins-builtin/        # 随包发布的插件（serial-bridge · rag-bridge · diag-bridge · twin-injection-default）
 │   └── plugins/                # 运行时装配（单例）
 ├── mcp/                        # aw mcp —— 零依赖 stdio MCP 网关（38 个工具，实例自动发现）

package/README.md

@@ -28,7 +28,7 @@
 
 **[简体中文](./README-zh.md)** · **[Documentation](https://kingdol666.github.io/AgentWorkShop)** · **[Releases](https://github.com/kingdol666/AgentWorkShop/releases)** · **[Changelog](./changelog.md)** · **[Plugin API](./docs/plugins.md)** · **[SDK](./docs/sdk.md)**
 
-<sub><b>v0.7.63</b> · 14 engines · 6 field protocols (5 built-in + serial plugin) + MES REST integration · data-analysis ledger with line-scoped permissions · 119 runtime settings · bilingual docs (简体中文 / English)</sub>
+<sub><b>v0.7.66</b> · 14 engines · 6 field protocols (5 built-in + serial plugin) + MES REST integration (mirror & live) · data-analysis ledger with line-scoped permissions · per-call tool parameter/binding preflight · 82 host tools · 119 runtime settings · bilingual docs (简体中文 / English)</sub>
 
 <br />
 
@@ -155,7 +155,7 @@ Recorded against a running instance: real DAQ history, real write control, real
 | **Channel group chat & native HITL** | The channel timeline is a real group chat: a member's request becomes a **trackable job** (not a lost message), and engines with a native ask surface (e.g. `omp`) route HITL questions through the platform — approve once, the engine continues with the receipt. Member permissions and notifications are first-class. |
 | **Scheduled tasks** | Any channel task can run on a **schedule** — fixed `interval` (60 s floor) or `daily` at a set time — managed from `/workshop/schedules` or the REST API, with per-run history, busy-guard (no overlapping runs of the same schedule) and channel-scoped visibility. |
 | **Persistent memory & channel process memory** | Private + channel-shared domains; FTS5 with CJK segmentation, optional vector hybrid recall, token-budgeted injection. On top sit **process memories**: deterministic root/child/lead event contracts, a canonical root summary, and a durable outbox (dead-letter + compensation worker) so the story of a goal survives restarts. |
-| **Four entry points** | One manager behind every door: **WS** (AEP v1 event stream with seq-resume), **MCP** (~25 in-process tools), **A2A** (JSON-RPC 2.0 + AgentCard), **REST**. |
+| **Four entry points** | One manager behind every door: **WS** (AEP v1 event stream with seq-resume), **MCP** (38 in-process tools), **A2A** (JSON-RPC 2.0 + AgentCard), **REST**. |
 
 #### Industrial stack
 
@@ -168,9 +168,10 @@ Recorded against a running instance: real DAQ history, real write control, real
 | **Recipe versioning & governance** | Parameter changes are versioned with attribution (user/agent/system + operator + reason). Roll back to any revision or the last-good batch — non-destructively. Stale-node params are skipped and clearly marked. |
 | **Recipe-chain optimization loop** | Agents optimize through recipe management, the way human process engineers do: `recipe_trial` batch-dispatches multi-parameter candidates **without writing a version** → re-measure → progress adopts via `recipe_update` (same recipe id, version +1) → regression triggers a **unified rollback** (`recipe_rollback` with `dispatch: true` = definition revert + batch restore to the PLC). Single-knob direct dispatch is out of the loop; trials run under four-layer bounds and a per-line cadence. |
 | **Channel-bound lines (read-only grant)** | A channel can bind a line: every member reads that line's ops logs, recipe history and live status without any node binding (tool outputs are tagged "channel-bound, read-only"), while writes still require node authorization. The line brief is injected into member prompts; binding changes hot-recycle member runtimes. |
-| **Data-analysis ledger (v0.7.62)** | Agents persist analysis results as first-class records: native `analysis_save` / `analysis_query` host tools write a dedicated `analysis_reports` table keyed by line + recipe + time window, with conclusions, metrics, data sources and actor attribution. The ledger is append-only for agents; humans revise via REST with editor attribution. A dedicated `/analysis` page streams new reports live (ops-log WS), filters by line / recipe / time / keyword, and edit/delete are permission-gated (`operate` line mode; admin-only delete). External analysis plugins integrate through the same ledger — `diag_run` runs IDD analyses in the background (`stability-report` for line health reports, `deep-diagnosis` for root cause), returns a `task_id` immediately, tracks progress via `diag_status`, and the finished report lands in the ledger via `analysis_save`. |
+| **Data-analysis ledger (v0.7.62)** | Agents persist analysis results as first-class records: native `analysis_save` / `analysis_query` host tools write a dedicated `analysis_reports` table keyed by line + recipe + time window, with conclusions, metrics, data sources and actor attribution. The ledger is append-only for agents; humans revise via REST with editor attribution. A dedicated `/analysis` page streams new reports live (ops-log WS), filters by line / recipe / time / keyword, and edit/delete are permission-gated (`operate` line mode; admin-only delete). External analysis plugins integrate through the same ledger — `diag_run` runs IDD analyses in the background across **six intents** (deep root-cause diagnosis, line stability report, quick screening, DOE analysis, tuning-experience mining, optimization campaign — scenario prompts injected per intent), returns a `task_id` immediately, tracks progress via `diag_status`, and the finished report lands in the ledger via `analysis_save`. |
 | **Semantic parameter surface & tuning loop** | Every dispatchable write converges on the **recipe face** (`recipe_update` / `recipe_apply` / `recipe_trial`, each gated by binding → HITL reason → active-run). Writes stay narrowed by the **four-layer bound** — node safe range ∩ parameter baseline bounds ∩ active product bounds ∩ active recipe window — and each adjustment opens a tuning record that `dcw_judge` must close (keep / rollback / uncertain), with `dcw_rollback` executing the undo. (v0.7.63 removed the legacy `param_control` / `param_read` / `dcw_read` direct-node tools — dead surface under permission model v2 — so the agent tool list stays clean.) |
 | **Line operations** | Lines → products → recipes → batch runs. Recipe windows gate acquisition and interlock writes; every sample is tagged `product/recipe/run` for per-batch isolation. |
+| **MES REST integration (two data paths)** | Production-management data joins the same governed workspace: a `mes-rest` node maps the MES API's catalog / current / history endpoints through **declarative request maps** (auth via `secretRef` → env token, never plaintext; hooks can post-process ingested data). Agents read it two ways — **live**: `mes_fetch` issues the mapped request and returns the API response (large windows paginate async into a `mesds-*` CSV dataset read via `mes_datasets` / `mes_dataset_read`; raw rows never enter the prompt); **mirrored**: an `http` DAQ node polls the same API into the local TSDB so `daq_query` serves it like any native node — and keeps serving it when the API is down. `mes_catalog` is line-scoped and fail-closed. |
 | **Multi-modal DAQ frame pipeline** | Multi-point profiles (thickness/scanner) and CCD image frames flow through template sink pipelines: vectors & metadata into Timescale (`daq_frames`), pixels into object storage (MinIO, auto disk fallback); derived-metric thresholds ride the existing alarm chain. |
 | **Production hardening (v0.7.58)** | Hold-heartbeat lifecycle fixed — failed writes never occupy the setpoint (no phantom re-dispatch), driver/connection/line changes clear the value, line stop suspends hold automatically; silent mock fallback blocked (unknown protocols fail loudly, plugin unload can't fake success); frame pixel sha256 fingerprints + day-directory object GC (DB cross-check before delete); Timescale daily chunks + columnstore compression (TSL); heap-watermark watch with crash/liveness/watermark auto-respawn; write-hold windows & trial cadence persisted across restarts; runtime grant re-checks include the channel-bound line (revocation takes effect everywhere). |
 | **Agent self-audit tools** | `line_context`, `ops_log`, `recipe_log`, `recipe_versions`, `dcw_journal` — agents see exactly which line/product/recipe they control, who did what, and how every value changed. |
@@ -180,6 +181,7 @@ Recorded against a running instance: real DAQ history, real write control, real
 | Capability | Why it matters |
 |---|---|
 | **Line-level permissions** | Industrial data is gated **per production line** with three states (none / read-only / operate), enforced in the data plane — regular users see no line data until granted. |
+| **Per-call tool preflight (v0.7.66)** | Every industrial tool call is checked before its handler runs: parameter shape, plus reference membership for `node_id` / `line_id` / `recipe_id` / `dataset_id` and chained dataset/job/model/plan/scene ids (resolved through lineage). References must fall inside the caller's bound face (bound nodes ∪ channel-bound line read-only ∪ nodes on a bound recipe's line). Out-of-scope calls are rejected with the caller's bound face and a concrete next step; authorization denials are audited. |
 | **Full-operation audit log** | Every user / agent / system action lands in one queryable log; operators are attributed to "Channel/Member", distinct from users and the system. Streamed live over WS. |
 | **Team-scoped plugin switches** | Each team (channel) keeps an **independent plugin switch set** (`channel_plugins`): a disabled plugin's tools are not injected into that team's agents. Plugins themselves are hot-managed via `aw plugin` and the `/plugins` page. |
 | **Plugin extension API** | A self-contained directory under `plugins/<name>/` enhances **both halves at once**: `index.mjs` (server: hooks, routes, agent tools, **DAQ read drivers / DCW write drivers**, frame processors, node templates, config groups, KV, timers) and `client.mjs` (browser: panels injected into named slots, i18n, settings UI). Three scopes — `builtin` (shipped) > `project` (checkout) > `user` (`~/.AgentWorkShop`) — with ~1 s hot reload on enable/disable **and on code edits**; a disabled plugin's drivers are removed on hot reload immediately. Built-in example: **serial-bridge** (serial communication: read/write drivers, serial probe API, frontend panel). Full contract in [`docs/plugins.md`](./docs/plugins.md). |
@@ -381,7 +383,7 @@ aw update --check                      # only report; nothing is installed
 npm install -g agentworkshop@latest    # manual equivalent
 ```
 
-Releases follow semver. `aw start` verifies the config root on every launch and migrates the legacy pre-`home` `data/` layout into it (newest file wins), so data survives upgrades. SQLite schema migrations run server-side at boot. Current version: **v0.7.64** — see [Releases](https://github.com/kingdol666/AgentWorkShop/releases).
+Releases follow semver. `aw start` verifies the config root on every launch and migrates the legacy pre-`home` `data/` layout into it (newest file wins), so data survives upgrades. SQLite schema migrations run server-side at boot. Current version: **v0.7.66** — see [Releases](https://github.com/kingdol666/AgentWorkShop/releases).
 
 ### Your first agent × line session (~2 minutes)
 
@@ -572,7 +574,7 @@ curl -X POST http://localhost:3000/api/workshop/schedules \
 | Entry | Endpoint | Audience |
 |---|---|---|
 | **WS** | `/api/workshop/ws?channelId=…` | Dashboards / UI — AEP v1 envelopes, per-channel monotonic `seq`, 5000-event ring, `lastSeq` resume, snapshot fallback. |
-| **MCP** | in-process server, ~25 tools | Agents (omp host tools) — management + job-face tools, channel-scoped. |
+| **MCP** | in-process server, 38 tools | Agents (omp host tools) — management + job-face tools, channel-scoped. |
 | **A2A** | `POST /api/workshop/a2a/:agentId/rpc` | External agents — JSON-RPC 2.0, `AgentCard` at `/card`, `tasks/sendSubscribe` SSE. |
 | **REST** | `/api/workshop/**` | Humans / scripts — full management face. |
 

package/app/stores/workshop/notifications.ts

@@ -171,11 +171,32 @@ export const useNotificationsStore = defineStore('workshop.notifications', () =>
   let inflightFetch: Promise<void> | null = null
   let lastAttemptAt = 0
   const FETCH_MIN_GAP_MS = 2000
+  // 熔断器:窗口内尝试数封顶,超限冷却静默丢弃(防上游紧循环把浏览器 socket 耗尽,
+  // 殃及同页全部请求 —— 实测 2.3 万次/15s 时 Chrome ERR_INSUFFICIENT_RESOURCES 风暴)
+  let breakerUntil = 0
+  let breakerWindowStart = 0
+  let breakerTries = 0
+  const BREAKER_WINDOW_MS = 10_000
+  const BREAKER_MAX_TRIES = 10
+  const BREAKER_COOLDOWN_MS = 30_000
   async function guardedFetch(fn: () => Promise<void>): Promise<void> {
     if (inflightFetch) return inflightFetch
-    const wait = FETCH_MIN_GAP_MS - (Date.now() - lastAttemptAt)
+    const now0 = Date.now()
+    if (now0 < breakerUntil) return
+    const wait = FETCH_MIN_GAP_MS - (now0 - lastAttemptAt)
     if (wait > 0) await new Promise(r => setTimeout(r, wait))
+    // 唤醒后复查:等待窗内的并发调用方都通过了入口检查,不复查会把整批一起放行
+    if (inflightFetch) return inflightFetch
     lastAttemptAt = Date.now()
+    if (lastAttemptAt - breakerWindowStart > BREAKER_WINDOW_MS) {
+      breakerWindowStart = lastAttemptAt
+      breakerTries = 0
+    }
+    breakerTries += 1
+    if (breakerTries > BREAKER_MAX_TRIES) {
+      breakerUntil = lastAttemptAt + BREAKER_COOLDOWN_MS
+      breakerTries = 0
+    }
     inflightFetch = (async () => {
       try {
         await fn()

package/config.yml

@@ -12,7 +12,7 @@
 app:
   name: AgentWorkShop
   title: AgentWorkShop
-  version: 0.7.64
+  version: 0.7.66
   description: 基于 Nuxt 4 的配置驱动运行范式
 
 # 运行模式（dev/prod 实际由启动命令决定，此处仅作运行时展示）

package/docs/cli.en.md

@@ -62,7 +62,7 @@ pnpm tui             # the terminal workbench
 ### 3. Quick use via npx (no global install)
 
 ```bash
-npx agentworkshop start            # fetch and start (pin a version with: npx agentworkshop@0.7.64 start)
+npx agentworkshop start            # fetch and start (pin a version with: npx agentworkshop@0.7.66 start)
 npx agentworkshop doctor           # subcommands work as usual
 npx -p agentworkshop aw doctor     # use -p when you want the name aw
 ```
@@ -79,7 +79,7 @@ npx -p agentworkshop aw doctor     # use -p when you want the name aw
 | Package manager | `pnpm@11.9.0` (`packageManager`, needed only for source development) |
 | Module format | ESM (`type: module`) |
 | License | `Apache-2.0` |
-| Current version | `0.7.64` (`package.json`) |
+| Current version | `0.7.66` (`package.json`) |
 
 ## 2. First start (the Claude Code experience)
 

package/docs/cli.md

@@ -60,7 +60,7 @@ pnpm tui             # 终端工作台
 ### 3. npx 快速使用（免全局安装）
 
 ```bash
-npx agentworkshop start            # 拉取并启动(可锁定版本: npx agentworkshop@0.7.64 start)
+npx agentworkshop start            # 拉取并启动(可锁定版本: npx agentworkshop@0.7.66 start)
 npx agentworkshop doctor           # 子命令照常可用
 npx -p agentworkshop aw doctor     # 需要 aw 这个名字时用 -p 指定包
 ```
@@ -76,7 +76,7 @@ npx -p agentworkshop aw doctor     # 需要 aw 这个名字时用 -p 指定包
 | 包管理器 | `pnpm@11.9.0`（`packageManager`，仅源码开发需要） |
 | 模块形态 | ESM（`type: module`） |
 | 许可证 | `Apache-2.0` |
-| 当前版本 | `0.7.64`（`package.json`） |
+| 当前版本 | `0.7.66`（`package.json`） |
 
 ## 二、首次启动（Claude Code 式）
 

package/docs/full-test-plan.md

@@ -1,6 +1,6 @@
 # AgentWorkShop 全功能测试流程 Plan
 
-> 版本无关的常备测试手册（撰写时基线 v0.7.63）。用途：后续直接指令 agent
+> 版本无关的常备测试手册（撰写时基线 v0.7.64）。用途：后续直接指令 agent
 > 「按 docs/full-test-plan.md 执行 Phase 0–N」即可完成全功能测试。
 > 断言数为历史全绿基线（2026-09 上旬各轮验收），以脚本实际输出为准；括号内数字仅用于识别明显回退。
 > 最后核对：2026-09-12（脚本清单逐一经 `scripts/` 实存核对）。

package/docs/plugins.en.md

@@ -9,7 +9,7 @@
 >
 > Applies to: plugin system v2 has been available since **v0.7.29** (frontend component
 > injection / plugin settings / plugin i18n / runtime services surface); this document
-> describes **v0.7.64**. `ctx.sdkVersion` is currently `0.3.0`.
+> describes **v0.7.66**. `ctx.sdkVersion` is currently `0.3.0`.
 
 ## 1. Folder layout & quick start
 

package/docs/plugins.md

@@ -6,7 +6,7 @@
 > 本文是唯一权威参考;`docs/site/plugins/` 的分页文档与本文同源。
 >
 > 适用版本:插件系统 v2 自 **v0.7.29** 起可用(前端组件注入 / 插件设置 / 插件 i18n /
-> 运行时服务面);本文档对应当前版本 **v0.7.64**。`ctx.sdkVersion` 当前为 `0.3.0`。
+> 运行时服务面);本文档对应当前版本 **v0.7.66**。`ctx.sdkVersion` 当前为 `0.3.0`。
 
 ## 一、目录结构与快速开始
 

package/docs/sdk.en.md

@@ -7,7 +7,7 @@
 
 **SDK version constants**: server-side `SDK_VERSION = '0.3.0'`, browser-side `CLIENT_SDK_VERSION = '0.3.0'`.
 These two constants are the **interface version of the SDK itself** and evolve **independently**
-from the npm package version (the `version` field in `package.json`, currently 0.7.64) —
+from the npm package version (the `version` field in `package.json`, currently 0.7.66) —
 them being out of sync is expected. Never use `SDK_VERSION` to infer the package version, or the reverse.
 **Module format**: ESM only (only `import` / `export`; there is no CJS entry).
 **Types**: `sdk/index.d.mts` and `sdk/client.d.mts` ship with the package, so TypeScript projects

package/docs/sdk.md

@@ -5,7 +5,7 @@
 > 零第三方运行时依赖,Node ≥ 23.4 与现代浏览器双端可用。
 
 **SDK 版本常量**:服务端 `SDK_VERSION = '0.3.0'`,浏览器端 `CLIENT_SDK_VERSION = '0.3.0'`。
-这两个常量是 **SDK 自身接口的版本号**,与 npm 包版本(`package.json` 的 `version`,当前 0.7.64)
+这两个常量是 **SDK 自身接口的版本号**,与 npm 包版本(`package.json` 的 `version`,当前 0.7.66)
 **各自独立演进** —— 常数不同步是预期行为,不要拿 `SDK_VERSION` 去推断包版本,反之亦然。
 **模块形态**:ESM only(只有 `import` / `export`,没有 CJS 入口)。
 **类型**:`sdk/index.d.mts`、`sdk/client.d.mts` 随包发布,TS 项目零配置获得提示(覆盖范围见 §7)。

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "agentworkshop",
-  "version": "0.7.64",
+  "version": "0.7.66",
   "type": "module",
   "description": "AgentWorkShop 软件开发系统 - 基于 Nuxt 4 的配置驱动运行范式",
   "license": "Apache-2.0",