Merge remote-tracking branch 'origin/main' into refactor/issue-fix-metrics-supplement-vocabulary-single-source

Signed-off-by: gcl-coder <88359731+gcl-coder@users.noreply.github.com>

# Conflicts:
#	examples/semantic-vocabulary-drift-smoke.py
#	loopx/semantics/vocabulary_v0.json

Author: gcl-coder

.github/workflows/full-public-smokes.yml

@@ -70,6 +70,9 @@ jobs:
         with:
           node-version: "24"
 
+      - name: Install the shell used by launch-agent contract checks
+        run: sudo apt-get update && sudo apt-get install -y zsh
+
       - name: Install smoke runtime dependencies
         run: python -m pip install --disable-pip-version-check "jsonschema>=4.23,<5"
 

README.md

@@ -9,6 +9,8 @@ The open, local-first control plane for long-horizon agents and personal agent t
 
 [![License](https://img.shields.io/badge/license-Apache--2.0-blue.svg)](LICENSE) [![Release](https://img.shields.io/github/v/release/loopx-project/loopx?filter=v*&display_name=tag)](https://github.com/loopx-project/loopx/releases/latest) [![Discord](https://img.shields.io/badge/Discord-Join-5865F2?logo=discord&logoColor=white)](https://discord.gg/XmGgQyCFZd)
 
+<a href="https://trendshift.io/repositories/102379"><img src="https://trendshift.io/api/badge/repositories/102379" alt="loopx-project/loopx on Trendshift" width="220" height="48"></a>
+
 [Get started](#try-loopx) · [Workspace](#meet-the-personal-agent-workspace) · [LHTB results](#lhtb-results) · [Docs](https://loopx-project.github.io/loopx/docs/) · [简体中文](README.zh-CN.md)
 
 **[LHTB](https://zli12321.github.io/LHTB/index.html) · 46 tasks · GPT-5.6 Sol:** LoopX 1.0.3 Heartbeat reaches **0.4948 mean reward** — **+17.3% vs Plain Codex**, **+10.6% vs native Codex Goal**.<br>
@@ -823,8 +825,6 @@ coordination, and an explicitly staged architecture and research incubator.
 
 ## Star History
 
-<a href="https://trendshift.io/repositories/102379"><img src="https://trendshift.io/api/badge/repositories/102379" alt="loopx-project/loopx on Trendshift" width="220" height="48"></a>
-
 <p align="center">
   <a href="https://github.com/loopx-project/loopx/stargazers"><img src="https://loopx-project.github.io/loopx/site-assets/star-history.svg" alt="LoopX GitHub star history from verified snapshots" width="800"></a><br>
   <sub>Generated every six hours from GitHub's official stargazer timestamps using a repository-authorized workflow. A snapshot is published only when the fetched rows match GitHub's current star count; GitHub's image cache may delay refreshes.</sub>

README.zh-CN.md

@@ -9,6 +9,8 @@
 
 [![License](https://img.shields.io/badge/license-Apache--2.0-blue.svg)](LICENSE) [![Release](https://img.shields.io/github/v/release/loopx-project/loopx?filter=v*&display_name=tag)](https://github.com/loopx-project/loopx/releases/latest) [![Discord](https://img.shields.io/badge/Discord-Join-5865F2?logo=discord&logoColor=white)](https://discord.gg/XmGgQyCFZd)
 
+<a href="https://trendshift.io/repositories/102379"><img src="https://trendshift.io/api/badge/repositories/102379" alt="loopx-project/loopx 在 Trendshift 的趋势排名" width="220" height="48"></a>
+
 [开始使用](#试用-loopx) · [个人工作区](#认识个人-agent-工作区) · [LHTB 结果](#lhtb-results) · [文档](https://loopx-project.github.io/loopx/docs/) · [English](README.md)
 
 **[LHTB](https://zli12321.github.io/LHTB/index.html) · 46 个任务 · GPT-5.6 Sol：**LoopX 1.0.3 Heartbeat 平均 Reward 达到 **0.4948**，较 **Plain Codex 提升 17.3%**，较**原生 Codex Goal 提升 10.6%**。<br>
@@ -736,8 +738,6 @@ benchmark 证据、operator surface 与 IM integration、shared-goal 跨 host 
 
 ## Star 趋势
 
-<a href="https://trendshift.io/repositories/102379"><img src="https://trendshift.io/api/badge/repositories/102379" alt="loopx-project/loopx 在 Trendshift 的趋势排名" width="220" height="48"></a>
-
 <p align="center">
   <a href="https://github.com/loopx-project/loopx/stargazers"><img src="https://loopx-project.github.io/loopx/site-assets/star-history.svg" alt="LoopX GitHub Star 历史趋势，来自已校验快照" width="800"></a><br>
   <sub>由仓库授权的 workflow 每 6 小时基于 GitHub 官方 stargazer 时间戳生成；仅当拉取条数与 GitHub 当前 Star 总数一致时发布。GitHub 图片缓存可能延迟刷新。</sub>

apps/presentation/dashboard/README.md

@@ -452,18 +452,6 @@ usable. It uses an installed Playwright package or the Codex bundled runtime
 when available, and starts Vite through the local `vite` package rather than
 depending on `npm` / `npx` being on `PATH`.
 
-The ops decision-freshness smoke protects the detailed `?view=ops` panel with
-two public fixtures: a live-like zero-item summary and a stale/rebase-required
-decision example. It verifies the rendered Chinese/English operator copy,
-counts, top affected goal, and exact-replay wording instead of relying only on
-source-string checks.
-
-The promotion-readiness smoke protects the detailed `?view=ops` panel with
-fresh, stale, and missing readiness fixtures. It verifies the status badges,
-readiness/rerun decision, artifact window, age, reason, and source-of-truth copy
-for canary promotion readiness. The canonical fixture/browser script is
-`examples/dashboard-promotion-readiness-browser-smoke.mjs`; use the npm script
-above instead of calling ad hoc duplicate filenames.
 The grouped demo-readiness path also runs `examples/promotion-gate-smoke.py`
 before browser checks, so the structured `gate_state`, `can_promote`, and
 `should_warn` contract is covered even when browser smokes are skipped.

apps/presentation/dashboard/src/data/chat-model.ts

@@ -83,6 +83,8 @@ export type CollaborationReadback = {
   schema_version: "collaboration_request_readback_v0";
   request_id: string;
   agent_id: string;
+  goal_id?: string;
+  decision_reason?: string;
   brief: {
     purpose: string;
     context: string;

apps/presentation/dashboard/src/data/status.ts

@@ -131,6 +131,8 @@ export const todoIndexSchema = z.object({
   current_projected_count: z.number().optional().default(0),
   rollout_event_count: z.number().optional().default(0),
   item_limit: z.number().optional().nullable(),
+  complete: z.boolean().optional(),
+  unavailable_goal_ids: z.array(z.string()).optional(),
   items: z.array(todoIndexItemSchema).optional().default([]),
 });
 
@@ -341,6 +343,19 @@ export const projectAssetTodoProjectionGapSchema = z.object({
   recommended_action: z.string().optional().nullable(),
 });
 
+export const nativeChildActivitySchema = z.object({
+  schema_version: z.literal("native_subagent_activity_v0"),
+  observation: z.enum(["unknown", "coordinator_reported"]),
+  host_attested: z.literal(false),
+  configured_limit: z.number().int().nonnegative(),
+  launched_count: z.number().int().nonnegative(),
+  skipped_count: z.number().int().nonnegative(),
+  capacity_rejected_count: z.number().int().nonnegative(),
+  host_failed_count: z.number().int().nonnegative(),
+  parent_accepted_count: z.number().int().nonnegative(),
+  turn_instance_id: z.string(),
+});
+
 export const projectAssetSchema = z.object({
   owner: z.string(),
   gate: z.string(),
@@ -351,6 +366,7 @@ export const projectAssetSchema = z.object({
   quota: quotaSchema.optional().nullable(),
   control_plane: controlPlaneSchema.optional().nullable(),
   orchestration: orchestrationPolicySchema.optional().nullable(),
+  native_child_activity: nativeChildActivitySchema.optional().nullable(),
   latest_validation: projectAssetLatestValidationSchema.optional().nullable(),
   stale_latest_run_warning: staleLatestRunWarningSchema.optional().nullable(),
   todo_projection_gap: projectAssetTodoProjectionGapSchema.optional().nullable(),

apps/presentation/dashboard/src/features/personal-workspace/collaboration-card.tsx

@@ -8,14 +8,16 @@ const copy = {
     acceptance: "验收要求", return: "需要回传", supplied: "已提供给接收方", pending: "等待接收方读取",
     decision: "接收方判断", unknown: "尚未记录", unavailable: "暂时无法读取",
     adopt: "已采纳", defer: "已暂缓", reject: "未采纳", no_change: "无需调整",
-    result: "结论已保存", delivered: "结论已回传", details: "查看交办内容",
+    result: "回复等待送达", delivered: "回复已送达", unverified: "回复送达尚未核验", details: "查看交办内容",
+    owner: "接收方", reason: "原因",
   },
   en: {
     title: "Delegation brief", context: "Context & corrections", constraints: "Constraints", inputs: "Inputs",
     acceptance: "Acceptance", return: "Expected return", supplied: "Supplied to receiver", pending: "Awaiting receiver read",
     decision: "Receiver decision", unknown: "Not recorded", unavailable: "Readback unavailable",
     adopt: "Adopted", defer: "Deferred", reject: "Rejected", no_change: "No change needed",
-    result: "Conclusion saved", delivered: "Conclusion returned", details: "View delegation details",
+    result: "Reply awaiting delivery", delivered: "Reply delivered", unverified: "Reply delivery unverified", details: "View delegation details",
+    owner: "Recipient", reason: "Reason",
   },
 };
 
@@ -27,12 +29,14 @@ export function CollaborationCard({ request }: { request?: CollaborationReadback
   const decision = request.decision as "adopt" | "defer" | "reject" | "no_change";
   const conclusion = request.returns.find((reply) => reply.phase === "conclusion");
   return <section className="personal-collaboration" aria-label={c.title}>
-    <header><strong>{brief.purpose}</strong><span>{request.agent_id}</span></header>
+    <header><strong>{brief.purpose}</strong><span>{c.owner}: {request.goal_id ? `${request.goal_id} / ` : ""}{request.agent_id}</span></header>
     <p className="personal-collaboration-status">
       <span>{request.read_status === "supplied" ? c.supplied : request.read_status === "unavailable" ? c.unavailable : c.pending}</span>
       <span>{c.decision}: {c[decision] ?? (request.decision === "unavailable" ? c.unavailable : c.unknown)}</span>
-      {conclusion ? <span>{conclusion.status === "delivered" ? c.delivered : c.result}</span> : null}
+      {conclusion ? <span>{conclusion.status === "delivered" ? c.delivered
+        : conclusion.status === "queued" ? c.result : c.unverified}</span> : null}
     </p>
+    {request.decision_reason ? <p>{c.reason}: {request.decision_reason}</p> : null}
     <details>
       <summary>{c.details}</summary>
       <h4>{c.context}</h4><p>{brief.context}</p>

apps/presentation/dashboard/src/features/personal-workspace/context-drawer.tsx

@@ -736,6 +736,18 @@ export function ContextDrawer({ agents, attentionHistory = [], onSelectAttention
               <button className="personal-secondary-action" onClick={() => callbacks.onRequestScheduleConfig?.("heartbeat", selection.item.goalId)} type="button"><Radio size={16} />{t("drawer.setupHeartbeat")}</button>
               <button className="personal-secondary-action" onClick={() => callbacks.onRequestScheduleConfig?.("monitor", selection.item.goalId)} type="button"><CalendarClock size={16} />{t("drawer.scheduleAdd")}</button>
             </div> : null}
+            {selection.item.nativeChildActivity?.observation === "coordinator_reported" ? (
+              <section className="personal-detail-card personal-native-child-activity">
+                <h3>{t("drawer.subagentReportTitle")}</h3>
+                <p>{t("drawer.subagentReportedActivity", {
+                  started: selection.item.nativeChildActivity.launched_count,
+                  skipped: selection.item.nativeChildActivity.skipped_count,
+                  rejected: selection.item.nativeChildActivity.capacity_rejected_count,
+                  failed: selection.item.nativeChildActivity.host_failed_count,
+                  accepted: selection.item.nativeChildActivity.parent_accepted_count,
+                })}</p>
+              </section>
+            ) : null}
             {selection.item.subagentExecution ? <section className="personal-detail-card personal-goal-subagents">
               <div className="personal-subagent-heading">
                 <div>

apps/presentation/dashboard/src/features/personal-workspace/i18n.tsx

@@ -259,6 +259,8 @@ const en = {
   "drawer.subagentConfirmEnable": "Confirm turning on sub-agent execution",
   "drawer.subagentCurrentBoundary": "Current task-domain restriction",
   "drawer.subagentDescription": "Allows the runtime to create temporary child agents for independent tasks only after Todo, quota, capability, and write-scope gates pass. It does not force parallel work or grant durable authority.",
+  "drawer.subagentReportTitle": "Child activity",
+  "drawer.subagentReportedActivity": "Latest coordinator report: {started} starts, {skipped} skips, {rejected} capacity rejections, {failed} host failures, {accepted} parent-accepted results. Host verification is unavailable.",
   "drawer.subagentDisable": "Preview turning off sub-agent execution",
   "drawer.subagentDisableSummary": "New child-agent execution will be disabled for this Goal. Existing Todo ownership and execution records stay unchanged.",
   "drawer.subagentDomainInvalid": "The selected task-domain restriction is invalid.",
@@ -1410,6 +1412,8 @@ const zhCN: Record<WorkspaceMessageKey, string> = {
   "drawer.subagentConfirmEnable": "确认开启子代理执行",
   "drawer.subagentCurrentBoundary": "当前任务领域限制",
   "drawer.subagentDescription": "仅在 Todo、配额、能力和写入范围门禁全部通过后，允许运行时为相互独立的任务临时创建子代理；不会强制并行，也不会授予持久权限。",
+  "drawer.subagentReportTitle": "子代理活动",
+  "drawer.subagentReportedActivity": "最近一轮主 Agent 回报：启动 {started} 次、跳过 {skipped} 次、容量拒绝 {rejected} 次、宿主失败 {failed} 次、主 Agent 验收 {accepted} 项；目前没有宿主核验。",
   "drawer.subagentDisable": "预览关闭子代理执行",
   "drawer.subagentDisableSummary": "这个 Goal 将不再创建新的子代理；现有 Todo 归属和执行记录不受影响。",
   "drawer.subagentDomainInvalid": "所选任务领域限制无效。",

apps/presentation/dashboard/src/features/personal-workspace/personal-workspace-model.ts

@@ -117,6 +117,16 @@ export type WorkspaceGoal = {
   repository?: WorkspaceRepositoryContext;
   state: WorkspaceGoalState;
   subagentExecution?: WorkspaceGoalSubagentConfiguration;
+  nativeChildActivity?: {
+    turn_instance_id: string;
+    observation: "unknown" | "coordinator_reported";
+    host_attested: false;
+    launched_count: number;
+    skipped_count: number;
+    capacity_rejected_count: number;
+    host_failed_count: number;
+    parent_accepted_count: number;
+  } | null;
   title: string;
   usage?: WorkspaceGoalUsage | null;
 };