@zq-silk/yui 1.1.2 → 2.1.0

package/skills/yui-runtime/SKILL.md

@@ -17,8 +17,8 @@ workspace layout, native transcript, or an earlier AgentRun.
 ## Enter through the current context
 
 There are two normal Task entry points. A user may continue directly with the
-current, unrevoked Leader Session: read current context using the Session CLI
-with `task context <task-id> --json`. Do not request a self-wake, reopen a Task,
+current, unrevoked Leader Session: read current context using
+`yui task context <task-id> --json`. Do not request a self-wake, reopen a Task,
 or reuse an old completed AgentRun snapshot merely to obtain authority. Pending
 delivery, unknown execution evidence and missing reports do not themselves
 revoke Session authority. Task lifecycle, scope, Assignment, workspace and
@@ -27,26 +27,35 @@ authority merely because the Task becomes active.
 Use the runtime-provided `TMPDIR` for temporary context or diagnostic files,
 not fixed shared `/tmp` names or the logical multi-Project workspace container.
 
+Ordinary commands use `yui` from the intended launch environment's PATH and
+`YUI_HOME`. Session/native identity, Task/Assignment/workspace authority and
+Controller protocol/storage checks still apply; they do not distinguish every
+installation with the same protocol and storage versions. If the entry or Home
+is wrong or unavailable, report that fact and use the explicitly authorized
+entry/environment; do not guess another installation or clear Session identity.
+
 For every explicitly dispatched managed Task AgentRun:
 
 1. Read the exact AgentRun identity from the newest Bootstrap Envelope.
-2. Before acting, load its authorized pack with the Session CLI named by the
-   current Session Manifest:
+2. Before acting, load its authorized pack using the current Session Manifest's
+   Context command and the exact Task/Run identity:
 
    ```sh
-   "$YUI_SESSION_CLI" task run context "$YUI_TASK_ID/<run-id>" --json
+   yui task run context "<task-id>/<run-id>" --json
    ```
 
 3. Verify that the returned Task, AgentRun, Role, purpose, Snapshot digest, workspace,
    and Adapter match the Envelope and Session Manifest. Stop and report a
    context-load failure if the pack is missing, stale, unauthorized, malformed,
    or mismatched. Never request an inline/full-prompt fallback.
-4. Use pack summaries and pointers first. Expand only an authorized ref when
+4. Use the pack's `pointers` (including each pointer's summary) first. They are
+   the single readable-ref inventory; `deltaRefs` contains only changed identities.
+   Expand only an authorized ref when
    its full value is needed, selecting it by the pointer's exact `store` and
    `refId`:
 
    ```sh
-   "$YUI_SESSION_CLI" task run context expand "$YUI_TASK_ID/<run-id>" <ref-id> --store <store> --mode full --json
+   yui task run context expand "<task-id>/<run-id>" <ref-id> --store <store> --mode full --json
    ```
 
    `--store` is required, even when the ref id is unique. Select both fields
@@ -119,14 +128,73 @@ TUI, with the configured Agent, permissions and workspace unchanged. A live
 unmanaged Session is not silently replaced or adopted; use an explicit Session
 lifecycle action before enabling controlled delivery.
 
-Context reads never consume queue entries. Read the referenced Message in full
-from Session Context. Native/transport acceptance is not implementation, and
+Context reads never consume queue entries. Global Context returns bounded
+`pending` and `recent` discovery pages, not Message bodies. Read a referenced
+original with `role message show <role> <message-id> --json`. Follow a pending
+page's `nextCursor` using `role message list <role> --pending --cursor <cursor>`
+when it is incomplete; accepted delivery does not erase the Message from the
+separate all-message list. Native/transport acceptance is not implementation, and
 `interrupt-requested` is not a stopped Turn or stopped background resources.
+Saving a Task Message or Brief is not notification delivery to Operator. A
+Task Role uses the existing Task InputRequest path for a genuinely missing
+user decision, not global Role input controls or fabricated progress questions.
 Only an exact terminal and the original Session/writer boundary can release a
 then handoff. An accepted or unconfirmed steer must not be submitted again by
 changing its request id or composing then. A conclusive rejection permits an
 explicit new control attempt; uncertainty does not.
 
+## Read bounded information completely when required
+
+Structured CLI results live in the top-level `data`, not a JSON-encoded
+`output` string. Task current Context includes bounded current facts and
+message references; it does not include event or terminal Run history.
+`collections` and `attention` identify incomplete discovery. Use
+`task context list <task> --store <store> [--cursor <cursor>]` to discover one
+family, then `task context inspect <task> --store <store> --ref <refId>
+--digest <digest>` to read its exact current value. Common Task lists use the
+same `items`, `total`, `complete`, and `nextCursor` contract. Do not treat the
+first page or a summary as complete requirements.
+
+Long detail reads return `contentPage` instead of the ordinary value (under
+`data.context` for Run context/expand, `data.contextDelta` for Run Context delta). Repeat the same read with
+`--cursor <contentPage.nextCursor>` until `complete: true`. Concatenate `text`
+in offset order and parse the combined JSON once, retaining the same source
+and digest. A source-change error requires a fresh read, not mixing old and
+new chunks. Never use a read cursor to repeat a mutation.
+
+Read all pages of a relevant original requirement, result, or authorization
+before acting on it. For a wake, read its complete fixed window and the
+referenced original Messages; the wake itself contains pointers, not reports.
+For unrelated discovery, stop once the necessary evidence is found. Task
+event deltas retain their fixed upper bound and are history, not replacement
+current state. No read acknowledges implementation or consumes a queue.
+
+## Discover operations for the current caller
+
+Use `yui help` or `yui --json help [command ...]` through the authorized CLI
+entry for the current caller's command view. Help, examples and shell completion
+share the catalog projection; public terminals and unmanaged Agents share one
+public view. This offline view is only guidance, not an authenticated permission
+or a way to switch Roles. An incomplete or stale managed identity must not be
+cleared to obtain public execution authority.
+
+When capability discovery appears in that view, use `capability search` for
+templates and `capability describe <name> --task
+<task-id> --input '<json>'` for supported exact authorization assessments.
+No `access` field means no exact authorization conclusion was made: supply the
+required parameters and follow the domain's existing execution checks. Hidden
+operations are omitted; `requestable` includes the exact missing Grant bounds
+and existing request route; `authorized` remains subject to execution-time
+revalidation. Provider outages do not change this authorization observation.
+
+Discovery never sends requests, consumes Grants, starts plugins or approves
+effects. If the Task actually needs a requestable operation, use the existing
+Assignment Message to Leader, Leader InputRequest, or Operator path. Preserve
+the exact action, resource/plugin, digest, environment and phase. A response is
+not a Grant. Existing source-authorized Leader Grant operations and Operator
+Grant authority remain unchanged; never expand an authorization source or
+solicit real-resource validation merely because an operation is requestable.
+
 ## Preserve intent and authority
 
 An analysis, diagnosis, or review request is read-only unless the user also

package/skills/yui-runtime/references/publication.md

@@ -87,7 +87,9 @@ is not Candidate acceptance, Review, Integration or Task completion.
 
 ## Archive separately
 
-Completion does not authorize archive. The Operator obtains authorization for
+Completion does not authorize archive. A delivery Leader with original user
+authorization may use [ordinary archive](../../yui-leader/references/authorization.md);
+this grants no force or abandonment authority. Otherwise the Operator obtains authorization for
 the exact Task, checks archive eligibility, then uses `--integrated` for verified
 merged delivery or `--abandon` for deliberate non-delivery. General archive
 approval never implies `--force` authority. Preserve the Task record.

package/dist/web/assets/client/taskSummary.js

@@ -1,350 +0,0 @@
-export const TASK_SUMMARY_SCRIPT = `
-import { node, clear } from "/assets/js/dom.js";
-import { richText, inputCard, observabilityMetricCard } from "/assets/js/components.js";
-import { formatDateTime } from "/assets/js/format.js";
-
-const sessionLabels = {
-  active: ["Recent activity", "近期有活动"], waiting: ["Waiting", "正在等待"],
-  quiet: ["Quiet · not a proven deadlock", "安静 · 尚不能判定卡死"],
-  diagnostic: ["Needs diagnosis", "需要诊断"], unknown: ["Observation unknown", "观测未知"],
-  stopped: ["Stopped / failed", "已停止／失败"], idle: ["Turn ended", "该轮已结束"],
-  background: ["Background unsettled", "后台未结清"]
-};
-const sayFor = locale => (en, cn) => locale.startsWith("zh") ? cn : en;
-function block(title, key) {
-  const section = node("section", "record-card summary-block");
-  section.dataset.summary = key;
-  section.append(node("h3", "", title));
-  return section;
-}
-function disclosure(title, key) {
-  const details = node("details", "record-card");
-  details.dataset.viewKey = key;
-  details.append(node("summary", "", title));
-  return details;
-}
-
-export function renderSessionSummary(container, observation, locale) {
-  clear(container);
-  const say = sayFor(locale);
-  container.append(node("h3", "", say("Session activity", "会话活动")));
-  if (!observation) {
-    container.append(node("p", "muted", say(
-      "Observation unavailable or not yet read; this does not mean zero running Sessions.",
-      "观测尚未读取或不可用；不代表运行会话为零。")));
-    return;
-  }
-  const metrics = node("div", "session-counts");
-  for (const [key, count] of Object.entries(observation.counts)) {
-    if (!count) continue;
-    const label = sessionLabels[key];
-    metrics.append(node("span", "session-count", count + " · " + say(...label)));
-  }
-  if (!observation.sessions.length) metrics.append(node("p", "muted", say(
-    "No selected native Session is recorded. Historical resources were not inspected.",
-    "未记录当前选用的原生会话；未盘点历史资源。")));
-  container.append(metrics, node("p", "muted", say(
-    "Scope: this Task's selected native Sessions. Activity is not delivery progress; no live process probe.",
-    "范围：本 Task 当前选用的原生会话。活动不等于交付进展；未探测实时进程。")));
-  container.append(node("small", "muted", say("Independent observation read: ", "独立观察读取于：")
-    + formatDateTime(observation.readAt, locale)));
-  const details = disclosure(say("Session facts and timestamps", "会话依据与时间"), "sessions");
-  for (const session of observation.sessions) {
-    const row = node("article", "session-row");
-    row.append(node("strong", "", session.roleName + " · " + say(...sessionLabels[session.group])));
-    row.append(node("p", "record-copy", session.reason));
-    if (session.waitingReason) row.append(node("p", "", say("Waiting for: ", "等待：") + session.waitingReason));
-    row.append(node("p", "muted", say("Last activity: ", "最近活动：")
-      + (session.lastActivityAt ? formatDateTime(session.lastActivityAt, locale) : say("unobserved", "未观测"))));
-    row.append(node("p", "muted", say("Native input updated: ", "原生输入记录更新：") + formatDateTime(session.sourceUpdatedAt, locale)));
-    if (session.operations.length) row.append(node("p", "muted", say(
-      "Operations without a terminal (not proof of recent activity): ",
-      "尚无终态的操作（不证明近期仍有活动）：") + session.operations.join(", ")));
-    if (session.background.length) row.append(node("p", "", say("Background: ", "后台：")
-      + session.background.map(item => item.id + " · " + item.execution).join(", ")));
-    row.append(node("small", "mono", [session.nativeSessionId, session.nativeTurnId, session.attemptId].filter(Boolean).join(" / ")));
-    details.append(row);
-  }
-  container.append(details);
-}
-
-export function renderTaskSummary(container, data, t, locale, actions, recordCard) {
-  const say = sayFor(locale);
-  const entries = store => data.core.records.filter(entry => entry.ref.store === store);
-  const available = store => entries(store).filter(entry => !entry.omitted).map(entry => entry.value);
-  const scope = node("p", "muted", say("Scope: ", "范围：") + data.task.id + " · "
-    + say("Reading does not accept results or acknowledge Messages.", "读取不验收成果，也不确认消息。"));
-  container.append(scope);
-  const grid = node("div", "task-summary-grid");
-  container.append(grid);
-  const attention = block(say("Needs your input", "需要你处理"), "attention");
-  attention.id = "detail-attention";
-  const inputCount = data.core.attention.openInputs.count;
-  attention.append(node("p", "", inputCount ? inputCount + say(" open questions", " 项待答问题")
-    : say("No open InputRequest requires your answer.", "当前没有待你回答的 InputRequest。")));
-  const inputs = entries("input-request").filter(entry => !entry.omitted && entry.value.status === "open");
-  inputs.forEach(entry => attention.append(inputCard(entry.value, { single: true }, t, locale, actions)));
-  const missing = data.core.attention.openInputs.refs.filter(ref => !inputs.some(entry => entry.ref.refId === ref.refId));
-  missing.forEach(ref => attention.append(recordCard({ ref, omitted: true })));
-  if (inputCount > inputs.length + missing.length) attention.append(node("p", "muted", say(
-    "Some open questions are outside this bounded read. Read all original questions with: ",
-    "部分待答问题未包含在本次有界读取中，请读取全部原问题：") + "yui task input list " + data.task.id));
-  const technical = node("div", "");
-  technical.dataset.technicalSummary = "";
-  attention.append(technical);
-  grid.append(attention);
-  const sessions = block("", "sessions");
-  sessions.dataset.sessionSummary = "";
-  grid.append(sessions);
-  const progress = block(say("Task progress", "任务进展"), "progress");
-  const briefEntry = entries("task-brief")[0];
-  const brief = available("task-brief")[0];
-  const working = data.task.status === "draft" || data.task.status === "active";
-  if (data.task.completionSummary) progress.append(richText(say("Recorded completion", "已记录完成"), data.task.completionSummary, t));
-  if (brief) {
-    progress.append(richText(working
-      ? say("Current focus", "当前工作")
-      : say("Last working focus", "结束前的工作重点"), brief.currentFocus, t));
-    progress.append(richText(working
-      ? say("Leader's progress report", "Leader 进展摘要")
-      : say("Last working report", "结束前的进展摘要"), brief.leaderSummary, t));
-    progress.append(node("small", "muted", say("Brief updated: ", "Brief 更新：") + formatDateTime(brief.updatedAt, locale)));
-  } else if (briefEntry) progress.append(recordCard(briefEntry));
-  else progress.append(node("p", "muted", say("No progress summary in this read; no percentage or ETA is inferred.",
-    "本次读取未包含进展摘要；不推算完成百分比或预计时间。")));
-  const next = node("p", "muted");
-  next.dataset.nextSummary = "";
-  progress.append(next);
-  grid.append(progress);
-  const conclusions = block(say("Key conclusions", "关键结论"), "conclusions");
-  const decisions = entries("task-decision").filter(entry => !entry.omitted && entry.value.status === "active");
-  decisions.slice(0, 3).forEach(entry => {
-    const card = recordCard(entry);
-    card.prepend(node("small", "muted", say("Recorded decision", "已记录决定")));
-    conclusions.append(card);
-  });
-  if (!decisions.length) conclusions.append(node("p", "muted", say(
-    "No current Decision is included. Reports and recommendations are not user approval.",
-    "本次读取未包含当前 Decision；报告和建议不代表用户批准。")));
-  if (decisions.length > 3 || data.core.omitted.records || entries("task-decision").some(entry => entry.omitted)) {
-    conclusions.append(node("p", "muted", say("More conclusions or omitted sources are available in history / Context.",
-      "更多结论或省略来源请查看历史／Context。")));
-  }
-  grid.append(conclusions);
-  updateTaskObservations(container, data, t, locale);
-}
-
-export function updateTaskObservations(container, data, t, locale) {
-  const say = sayFor(locale);
-  const sessions = container.querySelector("[data-session-summary]");
-  if (sessions && !sessions.contains(document.activeElement)) {
-    const wasOpen = sessions.querySelector("details")?.open;
-    renderSessionSummary(sessions, data.runtimeStatus === "available" ? data.runtime?.sessions : null, locale);
-    if (wasOpen && sessions.querySelector("details")) sessions.querySelector("details").open = true;
-  }
-  const technical = container.querySelector("[data-technical-summary]");
-  const execution = data.runtimeStatus === "available" ? data.runtime?.execution : null;
-  if (technical) {
-    clear(technical);
-    if (!execution) technical.append(node("p", "muted", say(
-      "Technical attention and native waits have not been observed; no open question does not prove health.",
-      "技术关注和原生等待尚未观测；没有待答问题不代表系统完全健康。")));
-    else {
-      const facts = execution.attention.concat(execution.blockers.filter(item => item.kind !== "input"));
-      if (facts.length) technical.append(node("h4", "", say("Technical attention · owner assigned, not necessarily started",
-        "技术关注 · 有负责人不代表已开始处理")));
-      facts.forEach(item => technical.append(node("p", "record-copy", item.owner + " · " + item.summary + " · " + item.id)));
-      if (!facts.length) technical.append(node("p", "muted", say("No technical attention in the current Task projection.",
-        "当前任务投影未列出技术关注。")));
-    }
-    const waits = data.runtimeStatus === "available" ? data.runtime?.sessions?.sessions.filter(session =>
-      session.group === "waiting" && ["user", "permission"].includes(session.waitingReason)) ?? [] : [];
-    waits.forEach(session => technical.append(node("p", "", session.roleName + say(
-      " is waiting for user/permission input. Inspect its original Session; this is not a new approval request.",
-      " 明确等待用户／权限输入。请查看原 Session；这里没有新建审批请求。"))));
-  }
-  const next = container.querySelector("[data-next-summary]");
-  if (next) next.textContent = execution ? say("Next owner: ", "下一步负责人：") + execution.next.owner
-    + " · " + t("exec.action." + execution.next.action) + " · " + execution.reason
-    : say("Next owner/action observation unavailable.", "下一步负责人／动作观测不可用。");
-}
-
-export function renderEvidence(container, data, t, locale, actions) {
-  const say = sayFor(locale);
-  const state = data.viewState;
-  const artifacts = disclosure(say("Saved files · read a fixed version", "已保存文件 · 读取固定版本"), "artifacts");
-  const body = node("div", "section-body");
-  artifacts.append(body);
-  const selection = node("div", "section-body");
-  let readSequence = 0;
-  const read = async (path, commit) => {
-    const sequence = ++readSequence;
-    selection.dataset.reading = "true";
-    selection.textContent = say("Reading selected revision…", "正在读取所选版本…");
-    try {
-      const result = await actions.readArtifact(data.task.id, path, commit);
-      if (sequence !== readSequence) return;
-      state.artifact = result;
-      drawSelection();
-    } catch (error) {
-      if (sequence === readSequence) selection.textContent = say("Selected revision unavailable: ", "所选版本不可用：") + error.message;
-    } finally { if (sequence === readSequence) selection.dataset.reading = "false"; }
-  };
-  function drawSelection() {
-    clear(selection);
-    const value = state.artifact;
-    if (!value) return;
-    const ref = "git:" + value.commit + ":" + value.relativePath;
-    selection.append(node("strong", "", say("Fixed result · ", "固定成果 · ") + value.relativePath));
-    selection.append(node("p", "mono", ref));
-    selection.append(node("p", "muted", say(
-      "This revision stays selected across refreshes. Reading or discussing it does not accept, publish or execute it.",
-      "刷新后仍保留此版本。读取或讨论不会验收、发布或执行该成果。")));
-    const copy = node("button", "record-open", say("Copy source for discussion", "复制来源用于讨论"));
-    copy.type = "button";
-    copy.addEventListener("click", async () => {
-      try { await navigator.clipboard.writeText(data.task.id + " " + ref); copy.textContent = say("Copied", "已复制"); }
-      catch { copy.textContent = say("Copy the source shown above", "请复制上方来源"); }
-    });
-    selection.append(copy);
-    const content = node("pre", "surface-json artifact-text", value.content);
-    content.tabIndex = 0;
-    content.setAttribute("aria-label", say("Fixed artifact text", "固定成果正文"));
-    selection.append(content);
-  }
-  const load = node("button", "record-open", say("Read current file list", "读取当前文件列表"));
-  load.type = "button";
-  const list = node("div", "section-body");
-  load.addEventListener("click", async () => {
-    load.disabled = true;
-    list.dataset.reading = "true";
-    try {
-      const result = await actions.artifacts(data.task.id);
-      state.artifactList = result;
-      drawList();
-    } catch (error) { list.textContent = say("File list unavailable: ", "文件列表不可用：") + error.message; }
-    finally { load.disabled = false; list.dataset.reading = "false"; }
-  });
-  function drawList() {
-    clear(list);
-    const current = state.artifactList;
-    if (!current) return;
-    if (state.artifact && current.commit !== state.artifact.commit) list.append(node("p", "", say(
-      "The file list is at a different revision. Your selected result below has not changed.",
-      "文件列表来自不同版本，下方所选成果未被替换。")));
-    if (!current.entries.length) list.append(node("p", "muted", say("No saved files at this read.", "本次读取没有已保存文件。")));
-    current.entries.forEach(entry => {
-      const button = node("button", "record-open artifact-choice", entry.relativePath + " · " + entry.size + " B");
-      button.type = "button";
-      button.addEventListener("click", () => read(entry.relativePath, current.commit));
-      list.append(button);
-    });
-  }
-  body.append(node("p", "muted", say("Files are loaded only on request; missing Context rows do not mean no files.",
-    "文件仅在请求时读取；Context 没有文件行不代表没有成果。")), load, list);
-  // Explicit frozen references stay attached to their original revision.
-  const sources = [data.task.completionArtifactRefs ?? [],
-    data.core.records.filter(entry => !entry.omitted && ["task-brief", "task-decision"].includes(entry.ref.store))
-      .map(entry => JSON.stringify(entry.value))].flat().join(" ");
-  const refs = [...sources.matchAll(/git:([0-9a-f]{40}):([^\\s"\\\\]+?)(?=[\\s"\\\\]|$)/g)];
-  const seen = new Set();
-  for (const match of refs) {
-    if (seen.has(match[0])) continue;
-    seen.add(match[0]);
-    const button = node("button", "record-open artifact-choice", say("Referenced revision · ", "已引用版本 · ") + match[2]);
-    button.type = "button";
-    button.addEventListener("click", () => read(match[2], match[1]));
-    body.append(button);
-  }
-  body.append(selection);
-  drawList();
-  drawSelection();
-  container.append(artifacts);
-  const delivery = disclosure(say("Delivery, review and retained resources", "交付、审阅与保留资源"), "delivery");
-  const output = node("div", "section-body");
-  delivery.append(output);
-  delivery.addEventListener("toggle", async () => {
-    if (!delivery.open) return;
-    clear(output);
-    const runtime = data.runtimeStatus === "available" ? data.runtime : null;
-    output.append(node("p", "", say("Local completion, review, remote delivery and resource cleanup are separate facts.",
-      "本地完成、审阅、远程交付和资源清理是不同事实。")));
-    if (!runtime) {
-      output.append(node("p", "muted", say("Evidence observation unavailable. Core requirements remain readable.",
-        "证据观测不可用；核心要求仍可读取。")));
-      return;
-    }
-    const remote = runtime.remoteDelivery;
-    if (remote) {
-      output.append(node("h4", "", say("Remote delivery", "远程交付") + " · " + remote.status));
-      remote.projects.forEach(project => {
-        const card = node("article", "record-card");
-        card.append(node("strong", "", project.directory + " · " + project.coverage));
-        card.append(node("p", "", project.reason));
-        card.append(node("p", "mono", say("Expected local commit: ", "期望本地提交：") + (project.expectedLocalCommit || say("unknown", "未知"))));
-        card.append(node("p", "", "PR/MR · " + (project.state || "none") + " · "
-          + say("Evidence: ", "证据：") + (project.verification || "unverified")));
-        if (project.adoption) card.append(node("p", "mono", say("Adoption: ", "采用记录：") + project.adoption.id));
-        output.append(card);
-      });
-    }
-    output.append(node("p", "muted", say(
-      "Review reports are evidence, not acceptance. Read the original report and its fixed candidate in Reviews below. Missing checks are not passes.",
-      "审阅报告是证据，不是验收。请在下方审阅中读取原报告与固定候选；缺失检查不算通过。")));
-    const original = node("div", "section-body");
-    output.append(original);
-    original.textContent = say("Reading original checks and review bindings…", "读取原始检查与审阅绑定…");
-    void actions.evidence(data.task.id).then(evidence => {
-      clear(original);
-      original.append(node("h4", "", say("Integration checks · exact recorded candidate", "集成检查 · 精确记录候选")));
-      if (!evidence.integrations.length) original.append(node("p", "muted", say(
-        "No Integration check records. Direct Leader verification may be recorded in the completion report.",
-        "没有 Integration 检查记录；Leader 直接执行的验证可能记录于完成报告。")));
-      for (const attempt of evidence.integrations) {
-        const card = node("article", "record-card");
-        card.append(node("strong", "", attempt.id + " · " + attempt.status),
-          node("p", "mono", (attempt.candidateCommit || say("No check candidate", "无检查候选"))),
-          node("p", "muted", say("Coverage of the physical current head has not been checked here.",
-            "此处未核对对物理当前 head 的覆盖。")));
-        if (!attempt.checks?.length) card.append(node("p", "muted", say("No check outcomes recorded.", "未记录检查结果。")));
-        (attempt.checks || []).forEach(check => card.append(node("p", "", check.name + " · " + check.outcome),
-          ...(check.details ? [node("p", "muted", check.details)] : [])));
-        original.append(card);
-      }
-      original.append(node("h4", "", say("Review bindings · not semantic approval", "审阅绑定 · 不等于语义通过")));
-      if (!evidence.reviews.length) original.append(node("p", "muted", say("No Review records.", "没有 Review 记录。")));
-      for (const round of evidence.reviews) {
-        const card = node("article", "record-card");
-        card.append(node("strong", "", round.id + " · " + round.status));
-        card.append(node("p", "mono", round.taskCandidate
-          ? round.taskCandidate.projects.map(p => p.projectId + " · " + p.commit).join("\\n")
-          : [round.workItemId, round.candidateId, round.reviewBaseCommit].filter(Boolean).join(" · ")));
-        if (round.reviewerRunId) {
-          const report = node("button", "record-open", say("Read original review report", "读取原始审阅报告"));
-          report.type = "button";
-          report.addEventListener("click", async () => {
-            report.disabled = true;
-            try {
-              const source = await actions.inspect(data.task.id, { store: "run", refId: round.reviewerRunId });
-              card.append(richText(round.reviewerRunId, source.value.result?.output
-                || say("No report recorded.", "未记录报告。"), t));
-            } catch (error) { card.append(node("p", "muted", error.message)); report.disabled = false; }
-          });
-          card.append(report);
-        }
-        original.append(card);
-      }
-      original.append(node("h4", "", say("Recorded workspace owners", "已记录工作区所有者")));
-      evidence.workspaces.forEach(workspace => original.append(node("p", "mono", workspace.owner.type + " · " + workspace.root)));
-    }).catch(error => { original.textContent = say("Original evidence unavailable: ", "原始证据不可用：") + error.message; });
-    const metrics = observabilityMetricCard(runtime.observability, t);
-    if (metrics) output.append(metrics);
-    output.append(node("h4", "", say("Retained workspaces · recorded ownership only", "保留工作区 · 仅已记录所有权")));
-    output.append(node("p", "muted", say(
-      "No physical cleanup inspection was run. Use the existing read-only preflight when needed; it does not authorize archive or cleanup:",
-      "未执行物理清理检查。需要时使用既有只读预检；预检不授权归档或清理：")));
-    output.append(node("code", "", "yui task archive-preflight " + data.task.id + " --integrated"));
-  });
-  container.append(delivery);
-}
-`;