agentworkshop 0.7.52 → 0.7.53

package/.output/server/package.json

@@ -1,6 +1,6 @@
 {
   "name": "agentworkshop-prod",
-  "version": "0.7.52",
+  "version": "0.7.53",
   "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>v0.7.52</b> · 14 个执行引擎 · 6 种现场协议（5 内置 + 串口插件） · 113 个运行时设置项 · 双语文档（简体中文 / English）</sub>
+<sub><b>v0.7.53</b> · 14 个执行引擎 · 6 种现场协议（5 内置 + 串口插件） · 115 个运行时设置项 · 双语文档（简体中文 / English）</sub>
 
 <br />
 
@@ -165,6 +165,8 @@ AgentWorkShop 起家于**多智能体软件工作坊**——Channel 内的编码
 | **人工审批的写控** | 数控下发经过「**安全量程 ∩ 活动配方窗口**」联锁 → 可选 **HITL 审批** → PLC 写入 → **回读校验** → 带签名的写历史。 |
 | **数控读写通道** | 每个控制节点都能沿**写入时所用的同一条标定链路**读回 PLC 当前值：周期读 + 按需读 + Agent 读取，SET 与 ACT 并排呈现——读是被动观测，永不被写联锁阻断。 |
 | **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。 |
 | **工艺参数语义映射与调控闭环** | Agent 用工程语义思考：`param_control(param, value)` 按**工艺参数**寻址（跨批次/换配方语义稳定），`param_read` 读回 PLC 当前值取证。每次写入被**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。 |
 | **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。 |
 | **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。 |
@@ -184,7 +186,7 @@ AgentWorkShop 起家于**多智能体软件工作坊**——Channel 内的编码
 | **运行时可观测** | `GET /api/system/monitor` 把 Agent 团队内部暴露成数字：根队列深度、看门狗介入次数、Harness 会话复用、记忆 outbox 积压——每项都有对应的文档化回退开关，新机制可以不重新部署就关掉。 |
 | **产线操作单页（`/operations`）** | 一页看全对产线的每一次触碰：实时操作流水（REST 快照 + WS 推流）把写控、数采、建模、配方、回退、报警、产线、人工、系统动作归入**九类分类**，写控行渲染结构化 **前值 → 后值 diff 徽章**（失败标红），并直接从事件流推断**闭环阶段**（监视 → 探索·小步 → 推荐搜索 → 影子验证 → 生产）。同一数据面也供工作台控制台的「产线作业」面板使用。 |
 | **外部 MCP 网关** | `aw mcp` 启动零依赖的 **stdio MCP 服务**（38 个工具），自动发现运行中的实例（环境变量 → 锁文件 → 配置文件），把产线、配方（含 mark-good）、受治理 DCW 写入、工艺参数语义映射、数采查询、频道、团队供给、模型晋升、判卷/回退与插件管理暴露给任意 MCP 客户端——受 `mcp.enabled` 运行时设置门控，附 `--doctor` 与 `--print-config` 辅助指令。 |
-| **全量配置驱动运行时** | 全部运行旋钮（记忆预算、上下文压缩、回退护栏、保留策略、备份、日志级别、系统时区、MCP 开关…）在设置描述符注册表声明一次，优先级 **config.yml < runtime-settings < env**——**113 个设置项、18 组**（34 live / 79 restart），代码零硬编码默认。 |
+| **全量配置驱动运行时** | 全部运行旋钮（记忆预算、上下文压缩、回退护栏、保留策略、备份、日志级别、系统时区、MCP 开关…）在设置描述符注册表声明一次，优先级 **config.yml < runtime-settings < env**——**113 个设置项、18 组**（36 live / 79 restart），代码零硬编码默认。 |
 | **可配置系统时区** | 一个 `time.timeZone` 设置项（IANA 时区，live 热生效）统一驱动人读时间——日志、聊天、审计、Agent 记录与定时任务；机器时间轴（数采样本、`timestamptz`）仍保存绝对时间。 |
 | **可配置节拍** | 采样与查询的默认值/下限全部是 **live 设置**（`daq.sampling.*`、`daq.query.*`）：热重载、create/patch 时钳制，Agent 工具描述实时携带当前值。 |
 
@@ -386,7 +388,7 @@ 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：**v0.7.52**——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：**v0.7.53**——见[版本发布](https://github.com/kingdol666/AgentWorkShop/releases)。
 
 ### 第一次「Agent × 产线」会话（约 2 分钟）
 
@@ -717,6 +719,8 @@ cd docs/site && npx vitepress build      # 生产构建 → .vitepress/dist
 | 六种现场协议：Modbus TCP · Modbus RTU-over-TCP · OPC UA · MQTT · HTTP + 内置 serial-bridge 串口插件（数采 + 数控双向） | 已交付 |
 | Harness 可用性探测 + 执行前引擎强校验（UI 禁选 + 409） | 已交付 |
 | Recipe 归因版本化 + 非破坏回退（界面 + Agent 工具） | 已交付 |
+| 配方链路优化闭环：recipe_trial 整批试验（不写版本）→ 进步才固化 → 统一回退（PLC 整批恢复） | 已交付 |
+| 频道绑定产线：全员只读日志/配方/状态 + 产线简报注入提示词 | 已交付 |
 | Agent 自查工具：line_context / ops_log / recipe_log / recipe_versions / dcw_journal | 已交付 |
 | 多 Harness 并行真实产线 live E2E（四引擎、一条产线） | 已交付 |
 | HITL 审批流经真实 OPC UA 写入验证 | 已交付 |

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>v0.7.52</b> · 14 engines · 6 field protocols (5 built-in + serial plugin) · 113 runtime settings · bilingual docs (简体中文 / English)</sub>
+<sub><b>v0.7.53</b> · 14 engines · 6 field protocols (5 built-in + serial plugin) · 115 runtime settings · bilingual docs (简体中文 / English)</sub>
 
 <br />
 
@@ -166,6 +166,8 @@ Recorded against a running instance: real DAQ history, real write control, real
 | **Human-approved write control** | DCW writes flow through **safe-range ∩ recipe-window** interlock → optional **HITL approval** → PLC write → **readback verification** → signed write history. |
 | **Read-write DCW channels** | Every control node also **reads its PLC value back** through the same calibration path it writes with: periodic + on-demand + agent reads surface **SET vs ACT** side by side — passive observation, never blocked by write interlocks. |
 | **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. |
 | **Semantic parameter mapping & tuning loop** | Agents reason in engineering semantics: `param_control(param, value)` addresses a **process parameter** (stable across batches and recipe swaps) instead of a raw node, and `param_read` reads back the PLC value for evidence. Every write is narrowed by a **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. |
 | **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. |
 | **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. |
@@ -185,7 +187,7 @@ Recorded against a running instance: real DAQ history, real write control, real
 | **Runtime observability** | `GET /api/system/monitor` exposes the agent-team internals as numbers: root queue depth, watchdog interventions, harness session reuse, pending memory outbox — each guarded by a documented rollback switch, so new machinery can be turned off without a redeploy. |
 | **Line operations feed (`/operations`)** | One page for every touch on the plant: a live operation feed (REST snapshot + WS stream) classifies writes, DAQ, AML, recipe, rollback, alarm, line, manual and system actions into **nine categories**, renders structured **before → after diff badges** on write rows (failures marked red), and infers the **closed-loop phase** (monitor → exploration · small-step → recommendation search → shadow verification → production) straight from the event stream. The same feed powers a line-operations panel inside the workshop console. |
 | **External MCP gateway** | `aw mcp` starts a zero-dependency **stdio MCP server** (38 tools) that discovers a running instance automatically (env → lock file → config file) and exposes lines, recipes (incl. mark-good), governed DCW writes, semantic parameter mapping, DAQ queries, channels, team provisioning, model promotion, judge/rollback and plugin management to any MCP client — gated by the `mcp.enabled` runtime setting, with `--doctor` and `--print-config` helpers. |
-| **Fully config-driven runtime** | Every runtime knob (memory budgets, compaction, rollback guardrails, retention, backups, log level, system timezone, MCP gate…) is declared once in the settings descriptor registry with precedence **config.yml < runtime-settings < env** — **113 settings across 18 groups** (34 live / 79 restart), no hardcoded defaults in code. |
+| **Fully config-driven runtime** | Every runtime knob (memory budgets, compaction, rollback guardrails, retention, backups, log level, system timezone, MCP gate…) is declared once in the settings descriptor registry with precedence **config.yml < runtime-settings < env** — **115 settings across 18 groups** (36 live / 79 restart), no hardcoded defaults in code. |
 | **Configurable system timezone** | One `time.timeZone` setting (IANA id, live-applied) drives human-facing timestamps — logs, chat, audit, agent records and schedules — through a single formatter; machine timelines (DAQ samples, `timestamptz`) stay absolute. |
 | **Configurable cadences** | Sampling and query defaults/floors are **live settings** (`daq.sampling.*`, `daq.query.*`): hot-reloaded, clamped on node create/patch, and agent tool descriptions always carry the current values. |
 
@@ -254,7 +256,7 @@ page headers stack, dense tables become scrollable ledgers with a pinned identit
 <td width="50%"><img src="https://raw.githubusercontent.com/kingdol666/AgentWorkShop/main/docs/readme-assets/shot-monitor.png" alt="Runtime monitor" width="100%" /><br/><sub><b>Runtime monitor.</b> Every wired channel, member count, dependency cycle and owner.</sub></td>
 </tr>
 <tr>
-<td><img src="https://raw.githubusercontent.com/kingdol666/AgentWorkShop/main/docs/readme-assets/shot-settings.png" alt="Settings" width="100%" /><br/><sub><b>Settings.</b> 113 keys across 18 groups, descriptor-driven — the same registry the CLI reads.</sub></td>
+<td><img src="https://raw.githubusercontent.com/kingdol666/AgentWorkShop/main/docs/readme-assets/shot-settings.png" alt="Settings" width="100%" /><br/><sub><b>Settings.</b> 115 keys across 18 groups, descriptor-driven — the same registry the CLI reads.</sub></td>
 <td><img src="https://raw.githubusercontent.com/kingdol666/AgentWorkShop/main/docs/readme-assets/shot-plugins.png" alt="Plugins" width="100%" /><br/><sub><b>Plugins.</b> Three scopes, hot reload on edit, per-team switches.</sub></td>
 </tr>
 </table>
@@ -377,7 +379,7 @@ 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: **v0.7.52** — 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: **v0.7.53** — see [Releases](https://github.com/kingdol666/AgentWorkShop/releases).
 
 ### Your first agent × line session (~2 minutes)
 
@@ -655,7 +657,7 @@ AgentWorkShop/
 ├── sdk/                        # agentworkshop/sdk — plugin context, hook bus, REST client, browser SDK
 ├── tui/                        # terminal workbench (aw tui)
 ├── shared/
-│   └── config/                 # schema.json (113 setting descriptors) + engine (merge/validate/persist) + path resolver
+│   └── config/                 # schema.json (115 setting descriptors) + engine (merge/validate/persist) + path resolver
 ├── config.yml                  # factory defaults (read at build/start; version comes from package.json)
 ├── .AgentWorkShop/             # config root in a checkout — prompts (versioned) + runtime overrides · data · logs · commands (git-ignored)
 ├── data/                       # legacy pre-migration location (auto-migrated into the config root)
@@ -727,6 +729,10 @@ building, so those files are the single source of truth for the single-page guid
 | Six field protocols: Modbus TCP · Modbus RTU-over-TCP · OPC UA · MQTT · HTTP + the built-in serial-bridge plugin (acquisition + write control) | Shipped |
 | Harness availability probing + dispatch-time engine checks (UI disable + 409) | Shipped |
 | Recipe versioning with attribution + non-destructive rollback (UI + agent tools) | Shipped |
+| Recipe-chain optimization loop: recipe_trial batch trials (no version) → adopt on progress → unified rollback with PLC restore | Shipped |
+| Channel-bound lines: read-only log/recipe/status access for all members, line brief injected into prompts | Shipped |
+| Recipe-chain optimization loop: recipe_trial batch trials (no version) → adopt on progress → unified rollback with PLC restore | Shipped |
+| Channel-bound lines: read-only log/recipe/status access for all members, line brief injected into prompts | Shipped |
 | Agent self-audit: line_context / ops_log / recipe_log / recipe_versions / dcw_journal | Shipped |
 | Multi-harness parallel live E2E on a real protocol line (four engines, one line) | Shipped |
 | HITL approval flow verified over real OPC UA writes | Shipped |

package/app/components/app-header/HitlBell.vue

@@ -61,7 +61,7 @@ const {
                 v-if="item.requestType || isQuestion(item)"
                 class="hitl-item-reqtype"
                 :data-type="requestTypeOf(item)"
-              >{{ isQuestion(item) ? '提问' : '审批' }}</span>
+              >{{ isQuestion(item) ? t('appHeader.hitlReqQuestion') : t('appHeader.hitlReqApproval') }}</span>
             </span>
             <span class="hitl-item-title">{{ item.title }}</span>
             <span
@@ -76,7 +76,7 @@ const {
               v-if="item.policy"
               class="hitl-item-policy"
             >
-              <span class="i-tabler-shield-lock" /> 策略:{{ policyLabel(item) }}
+              <span class="i-tabler-shield-lock" /> {{ t('appHeader.hitlPolicy', { p0: policyLabel(item) }) }}
             </span>
           </button>
 
@@ -85,19 +85,19 @@ const {
             v-if="!HITL_KINDS.includes(item.kind)"
             class="hitl-answer-note"
           >
-            该类型待办需在对应引擎界面处理
+            {{ t('appHeader.hitlEngineNote') }}
           </div>
           <div
             v-else-if="!permsKnown(item)"
             class="hitl-answer-note"
           >
-            读取审批权限…
+            {{ t('appHeader.hitlPermsLoading') }}
           </div>
           <div
             v-else-if="!canAnswer(item)"
             class="hitl-answer-note"
           >
-            无审批权限{{ item.policy ? `(${policyLabel(item)})` : '' }}
+            {{ t('appHeader.hitlNoPerm') }}{{ item.policy ? `(${policyLabel(item)})` : '' }}
           </div>
           <div
             v-else
@@ -142,7 +142,7 @@ const {
                   v-if="!(q.options && q.options.length > 0) || q.freeText"
                   v-model:value="qFree[slotOf(item, q.id)]"
                   :rows="2"
-                  :placeholder="q.options && q.options.length > 0 ? '补充说明(可选)' : '输入答案'"
+                  :placeholder="q.options && q.options.length > 0 ? t('appHeader.hitlSupplementPlaceholder') : t('appHeader.hitlAnswerPlaceholder')"
                 />
               </div>
               <div class="hitl-btns">
@@ -152,14 +152,14 @@ const {
                   :loading="hitl.answering === `${item.kind}:${item.id}`"
                   @click="submitQuestion(item)"
                 >
-                  提交答案
+                  {{ t('appHeader.hitlSubmitAnswer') }}
                 </a-button>
                 <a-button
                   size="small"
                   :disabled="hitl.answering === `${item.kind}:${item.id}`"
                   @click="submitCancel(item)"
                 >
-                  取消请求
+                  {{ t('appHeader.hitlCancelRequest') }}
                 </a-button>
               </div>
             </template>
@@ -202,7 +202,7 @@ const {
                   :disabled="hitl.answering === `${item.kind}:${item.id}`"
                   @click="submitCancel(item)"
                 >
-                  取消
+                  {{ t('common.cancel') }}
                 </a-button>
               </div>
             </template>

package/app/components/dcw/DcwAddNodeModal.vue

@@ -124,16 +124,16 @@ const addSemantics = defineModel<string>('semantics', { required: true })
           >
         </label>
         <label class="f">
-          <span>探索阶段单步上限 <em>(工程量)</em></span>
+          <span>{{ $t('dcwDetail.stepLimitLabel') }} <em>({{ $t('dcwDetail.stepLimitUnit') }})</em></span>
           <input
             v-model.number="addStepLimit"
             type="number"
             min="0"
             step="any"
             class="inp"
-            placeholder="留空：按节点量程 2% 默认值"
+            :placeholder="$t('dcwDetail.stepLimitPlaceholder')"
           >
-          <small class="field-hint">Agent 每次写入的最大变化量；无有限值时探索写入会被拒绝。</small>
+          <small class="field-hint">{{ $t('dcwDetail.stepLimitHint') }}</small>
         </label>
       </div>
 

package/app/components/dcw/DcwNodesTable.vue

@@ -54,7 +54,7 @@ function saveStep(n: DcwNodeView): void {
             <th>{{ $t('dcwDetail.k1dexou6058') }}</th>
             <th>{{ $t('dcwDetail.k1b1nnaa059') }}</th>
             <th>{{ $t('dcwDetail.k1i8rtqt060') }}</th>
-            <th>探索步长</th>
+            <th>{{ $t('dcwDetail.thStepLimit') }}</th>
             <th class="right">
               {{ $t('dcwDetail.k40aa6061') }}
             </th>
@@ -168,7 +168,7 @@ function saveStep(n: DcwNodeView): void {
                   step="any"
                   class="inp step-inp mono"
                   placeholder="2% default"
-                  title="在线控制单步最大变化量；留空将关闭探索写入"
+                  :title="$t('dcwDetail.stepInputTitle')"
                   @input="stepDrafts[n.id] = ($event.target as HTMLInputElement).value === '' ? '' : Number(($event.target as HTMLInputElement).value)"
                 >
                 <button
@@ -176,10 +176,10 @@ function saveStep(n: DcwNodeView): void {
                   :disabled="stepPatchingId === n.id"
                   @click="saveStep(n)"
                 >
-                  {{ stepPatchingId === n.id ? '…' : '保存' }}
+                  {{ stepPatchingId === n.id ? '…' : $t('common.save') }}
                 </button>
               </div>
-              <small class="dim">当前 Δ≤{{ n.stepLimit ?? '未配置' }} {{ n.unit }}</small>
+              <small class="dim">{{ $t('dcwDetail.stepCurrent', { p0: n.stepLimit ?? $t('dcwDetail.stepUnset') }) }} {{ n.unit }}</small>
             </td>
             <td class="right">
               <button

package/app/components/dcw/DcwRecipeEditModal.vue

@@ -131,14 +131,14 @@ const recipeForm = defineModel<DcwRecipeForm>('form', { required: true })
           >
         </label>
         <label class="f">
-          <span>Recipe 单步上限 Δ≤</span>
+          <span>{{ $t('dcwDetail.recipeStepLimitLabel') }}</span>
           <input
             v-model.number="p.stepLimit"
             type="number"
             min="0"
             class="inp"
             step="any"
-            placeholder="跟随节点默认"
+            :placeholder="$t('dcwDetail.recipeStepLimitPlaceholder')"
           >
         </label>
         <button

package/app/components/workshop/ChatMemberPanel.vue

@@ -16,6 +16,7 @@ import type { AepChannelMember } from '#shared/workshop-protocol'
 
 const props = defineProps<{ channelId: string }>()
 
+const { t } = useI18n()
 const chat = useChatStore()
 
 const perms = computed<ChatPermissions | null>(() => chat.permissions[props.channelId] ?? null)
@@ -78,14 +79,14 @@ async function onJoin(): Promise<void> {
   busy.value = 'join'
   try {
     const status = await chat.join(props.channelId)
-    if (status === 'pending') message.info('已提交加入申请,等待 owner 批准')
-    else message.success('已加入群聊')
+    if (status === 'pending') message.info(t('chat.joinedPending'))
+    else message.success(t('chat.joinedOk'))
     await refresh()
   }
   catch (e) {
     const code = chatErrorCode(e)
-    if (code === 'CHAT_DISABLED') message.warning('该 Channel 未开启群聊')
-    else if (code === 'CHANNEL_PRIVATE') message.warning('该 Channel 未公开,需 owner 邀请')
+    if (code === 'CHAT_DISABLED') message.warning(t('chat.errChatDisabled'))
+    else if (code === 'CHANNEL_PRIVATE') message.warning(t('chat.errChannelPrivate'))
     else message.error(chatErrorMessage(e))
   }
   finally {
@@ -98,11 +99,11 @@ async function onLeave(): Promise<void> {
   busy.value = 'leave'
   try {
     await chat.leave(props.channelId)
-    message.success('已退出群聊')
+    message.success(t('chat.leftOk'))
     await refresh()
   }
   catch (e) {
-    if (chatErrorCode(e) === 'OWNER_CANNOT_LEAVE') message.warning('owner 不能退出群聊,请先转移 owner')
+    if (chatErrorCode(e) === 'OWNER_CANNOT_LEAVE') message.warning(t('chat.errOwnerCannotLeave'))
     else message.error(chatErrorMessage(e))
   }
   finally {
@@ -115,7 +116,7 @@ async function onApprove(m: AepChannelMember): Promise<void> {
   busy.value = `approve:${m.userId}`
   try {
     await chat.approveMember(props.channelId, m.userId)
-    message.success('已批准加入')
+    message.success(t('chat.approvedOk'))
   }
   catch (e) {
     message.error(chatErrorMessage(e))
@@ -130,7 +131,7 @@ async function onRemove(m: AepChannelMember): Promise<void> {
   busy.value = `remove:${m.userId}`
   try {
     await chat.removeMember(props.channelId, m.userId)
-    message.success('已移除成员')
+    message.success(t('chat.removedOk'))
   }
   catch (e) {
     message.error(chatErrorMessage(e))
@@ -156,12 +157,12 @@ async function onSaveSettings(): Promise<void> {
       approvalPolicy: d.approvalPolicy,
       chatEnabled: d.chatEnabled ? 1 : 0,
     })
-    message.success('群聊设置已保存')
+    message.success(t('chat.settingsSaved'))
     syncDraft()
   }
   catch (e) {
     if (chatErrorCode(e) === 'VERSION_CONFLICT') {
-      settingsError.value = 'Channel 设置已被他人修改,请刷新后重试'
+      settingsError.value = t('chat.settingsConflict')
     }
     else if (chatErrorCode(e) === 'CHAT_DISABLED') {
       settingsError.value = chatErrorMessage(e)
@@ -175,12 +176,12 @@ async function onSaveSettings(): Promise<void> {
   }
 }
 
-const statusLabel: Record<AepChannelMember['status'], string> = {
-  active: '在群',
-  pending: '待批准',
-  left: '已退出',
-  removed: '已移除',
-}
+const statusLabels = computed<Record<AepChannelMember['status'], string>>(() => ({
+  active: t('chat.statusActive'),
+  pending: t('chat.statusPending'),
+  left: t('chat.statusLeft'),
+  removed: t('chat.statusRemoved'),
+}))
 
 const avatarText = (m: AepChannelMember): string =>
   (m.displayName || m.userId.slice(0, 8)).trim().charAt(0).toUpperCase()
@@ -190,13 +191,13 @@ const avatarText = (m: AepChannelMember): string =>
   <div class="member-panel">
     <div class="mp-head">
       <span class="i-tabler-users-group mp-icon" />
-      <span class="mp-title">成员</span>
+      <span class="mp-title">{{ $t('chat.title') }}</span>
       <span class="mp-count">{{ members.filter(m => m.status === 'active').length }}</span>
       <span class="mp-spacer" />
       <button
         type="button"
         class="mp-refresh"
-        title="刷新成员与设置"
+        :title="$t('chat.refreshTitle')"
         @click="refresh().then(syncDraft)"
       >
         <span class="i-tabler-refresh" />
@@ -215,23 +216,23 @@ const avatarText = (m: AepChannelMember): string =>
             :loading="busy === 'join'"
             @click="onJoin"
           >
-            加入群聊
+            {{ $t('chat.join') }}
           </a-button>
           <div
             v-else
             class="mp-hint"
           >
-            {{ perms === null ? '未加入群聊 —— 需 owner 邀请或 Channel 未公开' : '该 Channel 未公开或未开启群聊 —— 需 owner 邀请' }}
+            {{ perms === null ? $t('chat.notMemberHint') : $t('chat.privateHint') }}
           </div>
         </template>
         <template v-else-if="myStatus === 'pending'">
           <div class="mp-hint pending">
-            <span class="i-tabler-hourglass" /> 加入申请待 owner 批准
+            <span class="i-tabler-hourglass" /> {{ $t('chat.pendingHint') }}
           </div>
         </template>
         <template v-else-if="!isOwner">
           <a-popconfirm
-            title="退出群聊后将立即失去读、发言、审批与通知补拉权限,确认退出?"
+            :title="$t('chat.leaveConfirm')"
             :ok-text="$t('common.confirm')"
             :cancel-text="$t('common.cancel')"
             @confirm="onLeave"
@@ -241,7 +242,7 @@ const avatarText = (m: AepChannelMember): string =>
               block
               :loading="busy === 'leave'"
             >
-              退出群聊
+              {{ $t('chat.leave') }}
             </a-button>
           </a-popconfirm>
         </template>
@@ -249,7 +250,7 @@ const avatarText = (m: AepChannelMember): string =>
           v-else
           class="mp-hint"
         >
-          <span class="i-tabler-crown" /> 你是 owner(不可退出,可转移 owner 或删除 Channel)
+          <span class="i-tabler-crown" /> {{ $t('chat.ownerHint') }}
         </div>
       </div>
 
@@ -274,7 +275,7 @@ const avatarText = (m: AepChannelMember): string =>
               <span
                 class="mp-status"
                 :data-status="m.status"
-              >{{ statusLabel[m.status] }}</span>
+              >{{ statusLabels[m.status] }}</span>
               <span class="mp-joined">{{ formatLocalClock(m.joinedAt, false) }}</span>
             </div>
           </div>
@@ -290,11 +291,11 @@ const avatarText = (m: AepChannelMember): string =>
               :loading="busy === `approve:${m.userId}`"
               @click="onApprove(m)"
             >
-              批准
+              {{ $t('chat.approve') }}
             </a-button>
             <a-popconfirm
               v-if="m.status === 'active' || m.status === 'pending'"
-              title="移除该成员后其立即失去群聊与审批权限,确认?"
+              :title="$t('chat.removeConfirm')"
               :ok-text="$t('common.remove')"
               :cancel-text="$t('common.cancel')"
               @confirm="onRemove(m)"
@@ -313,7 +314,7 @@ const avatarText = (m: AepChannelMember): string =>
           v-if="sortedMembers.length === 0"
           class="mp-hint"
         >
-          名册不可见(非成员)
+          {{ $t('chat.rosterHidden') }}
         </div>
       </div>
 
@@ -323,35 +324,35 @@ const avatarText = (m: AepChannelMember): string =>
         class="mp-settings"
       >
         <div class="mp-sec-title">
-          群聊设置
+          {{ $t('chat.settingsTitle') }}
           <span class="mp-ver">v{{ settings?.version ?? '-' }}</span>
         </div>
         <label class="mp-field">
-          <span>可见性</span>
+          <span>{{ $t('chat.visibilityLabel') }}</span>
           <a-select
             v-model:value="draft.visibility"
             size="small"
-            :options="[{ value: 'private', label: '私有(仅成员可见)' }, { value: 'public', label: '公开(登录用户可发现)' }]"
+            :options="[{ value: 'private', label: $t('chat.visibilityPrivate') }, { value: 'public', label: $t('chat.visibilityPublic') }]"
           />
         </label>
         <label class="mp-field">
-          <span>加入策略</span>
+          <span>{{ $t('chat.joinPolicyLabel') }}</span>
           <a-select
             v-model:value="draft.joinPolicy"
             size="small"
-            :options="[{ value: 'owner_approve', label: '需 owner 批准' }, { value: 'open', label: '开放加入' }]"
+            :options="[{ value: 'owner_approve', label: $t('chat.joinPolicyOwnerApprove') }, { value: 'open', label: $t('chat.joinPolicyOpen') }]"
           />
         </label>
         <label class="mp-field">
-          <span>审批策略</span>
+          <span>{{ $t('chat.approvalPolicyLabel') }}</span>
           <a-select
             v-model:value="draft.approvalPolicy"
             size="small"
-            :options="[{ value: 'owner_only', label: '仅 owner 可决策' }, { value: 'any_member', label: '任一成员可决策' }]"
+            :options="[{ value: 'owner_only', label: $t('chat.approvalOwnerOnly') }, { value: 'any_member', label: $t('chat.approvalAnyMember') }]"
           />
         </label>
         <label class="mp-field switch">
-          <span>开启群聊</span>
+          <span>{{ $t('chat.chatEnabledLabel') }}</span>
           <a-switch
             v-model:checked="draft.chatEnabled"
             size="small"
@@ -367,7 +368,7 @@ const avatarText = (m: AepChannelMember): string =>
             type="link"
             @click="refresh().then(syncDraft)"
           >
-            刷新
+            {{ $t('chat.refresh') }}
           </a-button>
         </div>
         <a-button
@@ -377,7 +378,7 @@ const avatarText = (m: AepChannelMember): string =>
           :loading="busy === 'settings'"
           @click="onSaveSettings"
         >
-          保存设置
+          {{ $t('chat.saveSettings') }}
         </a-button>
       </div>
     </div>