@agile-team/wl-skills-ui 1.16.0 → 1.17.0

package/CHANGELOG.md

@@ -4,6 +4,11 @@ All notable changes to **@agile-team/wl-skills-ui** will be documented in this f
 
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/).
 
+## [1.17.0] - 2026-10-09
+
+- 统一可见 notice（包版本/判定/技能或基础约束/规则/目标/runId/待检查项）、开始与收尾提示、规范与执行器版本漂移诊断；保持单包使用与宿主证据未验证口径。
+- 按本包规则与实际 Profile 映射 Skill，不重复前端注释规则；保留旧入口哈希证明迁移与外来内容保护。
+
 ## [Unreleased]
 
 ## [1.16.0] - 2026-10-08

package/README.md

@@ -1,12 +1,20 @@
 # @agile-team/wl-skills-ui
 
-**企业级 UI 一致性与 AI 精确治理工具包 v1.16.0** — 面向 Vue 3 + Element Plus 的设计令牌、能力 Profile、样式皮肤、页面骨架、运行时能力、静态扫描、确定性修复、AI Skill、MCP 与脱敏 UI 模板契约。
+**企业级 UI 一致性与 AI 精确治理工具包 v1.17.0** — 面向 Vue 3 + Element Plus 的设计令牌、能力 Profile、样式皮肤、页面骨架、运行时能力、静态扫描、确定性修复、AI Skill、MCP 与脱敏 UI 模板契约。
 
 它把“视觉统一、问题识别、机械修复、项目适配和样例沉淀”拆成可验证、可组合的工程能力。脚本、规则和 MCP 先完成确定性工作，AI 只处理业务语义与人工决策，从而减少重复读源码、上下文 token 和误修改。
 
 [![npm version](https://img.shields.io/npm/v/@agile-team/wl-skills-ui.svg)](https://www.npmjs.com/package/@agile-team/wl-skills-ui)
 [![Node](https://img.shields.io/badge/node-%3E%3D22-green.svg)]()
 
+## 如何确认本包正在起作用
+
+每次适用任务先运行项目本地 `wl-ui task "修改目标文件" --target src/Foo.vue --json`。编辑前展示真实 `notice`：包名与版本、判定、选中 Skill 或基础约束、具体规则、目标、runId 和尚未执行的检查。普通修改也需要基础约束提醒；相关但未覆盖的意图显示 gap 与建议；没有目标依据显示 needs-context，职责外显示 not-applicable，不强行匹配。
+
+执行实际检查时复用 `--run-id`，结束读取 `wl-ui status --run-id <id> --json`，分别报告执行和验证、实际检查文件、过期证据与未执行项。`notice.displayEvidence=unverified` 表示工具回执不能证明聊天界面展示；安装、路由或模型自报不能证明宿主加载/模型读取。
+
+`wl-ui doctor-host --json` 对比已分发规范、本地执行器和正在运行的版本，漂移会显式报告。规范更新不会替代依赖升级：同步本包依赖、锁文件和受管入口；未使用的兄弟包无需安装。重开/刷新宿主加载后仍需观察真实任务调用，不能宣称所有 AI 自动触发。
+
 ## 你能获得什么
 
 | 目标 | 包提供的能力 | 带来的效果 |
@@ -500,7 +508,7 @@ npx wl-ui doctor --project . --print-overrides
 
 ## 版本与进一步阅读
 
-当前 v1.16.0：公开集成协议 describe/request 统一信封（能力清单与路由语料经真实打包产物验证）、输入校验收口与 MCP 目录对齐真实 tools/list。v1.15.0：任务判定、真实执行/验证回执、实际检查范围和过期检测；Codex 原生薄入口按需加载 canonical Skill。确定性路由覆盖已发布目录与显式缺口，不保证识别所有自然语言。v1.14.0：明确独立能力边界，安装、更新、清理按本包贡献保护用户及其他包内容；共享文件采用单文件原子替换，支持历史路由与 `.clinerules` 目录迁移。v1.13.2：R044 审查多列表单超宽标签，R040 识别自定义 `stepper`，均供人工复核。v1.13.0：新增 `native-jh-ag` Profile、显式 Profile 兼容承诺与联邦 AG 依赖识别。
+当前 v1.17.0：公开集成协议 describe/request 统一信封（能力清单与路由语料经真实打包产物验证）、输入校验收口与 MCP 目录对齐真实 tools/list。v1.15.0：任务判定、真实执行/验证回执、实际检查范围和过期检测；Codex 原生薄入口按需加载 canonical Skill。确定性路由覆盖已发布目录与显式缺口，不保证识别所有自然语言。v1.14.0：明确独立能力边界，安装、更新、清理按本包贡献保护用户及其他包内容；共享文件采用单文件原子替换，支持历史路由与 `.clinerules` 目录迁移。v1.13.2：R044 审查多列表单超宽标签，R040 识别自定义 `stepper`，均供人工复核。v1.13.0：新增 `native-jh-ag` Profile、显式 Profile 兼容承诺与联邦 AG 依赖识别。
 
 - [版本摘要](docs/version-history.md)
 - [完整 Changelog](CHANGELOG.md)

package/bin/task-gateway.mjs

@@ -9,7 +9,9 @@ description: Vue and Element Plus UI consistency, design tokens, styles, compone
 # wl-skills-ui task gateway
 For one user task reuse the same \`--run-id\` or \`WL_TASK_RUN_ID\` across installed applicable packages. Do not install unused sibling packages or create unrelated IDs.
 
-1. Start each task with \`node node_modules/@agile-team/wl-skills-ui/bin/wl-ui.js task "<task>" --target <path>\`. Retain the run ID and briefly explain the decision, applicable rules and unverified items. A routing decision does not execute checks.
+多项目工作区先沿目标路径向上找到本包安装清单与项目 AGENTS，切到该项目根再调用工具；不要以聚合工作区根代替 projectRoot，也不要在聚合根安装。不同项目分别保存项目身份，同一用户任务复用 runId。
+
+1. Start each relevant task from its project root with \`node node_modules/@agile-team/wl-skills-ui/bin/wl-ui.js task "<task>" --target <path> --json\`. Before editing, visibly show the returned \`notice\`: actual package/version, decision, Skill or baseline, rule IDs/names, target, runId and unverified checks. Include baseline and gaps. Report missing/unsupported local commands and version drift explicitly. A routing decision does not execute checks.
 2. Read \`node_modules/@agile-team/wl-skills-ui/SKILL.md\`, then only the selected canonical Skill under \`node_modules/@agile-team/wl-skills-ui/skills\`. Resolve references relative to that Skill. Honor the actual project Profile; do not infer a table migration from a Skill match.
 3. Run the actual scan/check/fix tools with the same \`--run-id\`. For a code gate use \`wl-ui scan --target <path> --fail-on-error --run-id <id>\`. An ordinary Vue edit still receives UI baseline checks; tasks outside the domain are not applicable.
 4. End with \`wl-ui status --run-id <id>\` and report execution separately from validation. Zero files, skipped checks, dependency-only checks and unresolved review items are not proof of all UI constraints passing.

package/bin/task-integration.mjs

@@ -41,7 +41,7 @@ export function taskCatalog() {
   return catalogFiles(join(root, "skills")).map((file) => {
     const id = relative(join(root, "skills"), file).replace(/\\/g, "/").replace(/\/SKILL\.md$/, "");
     const content = readFileSync(file, "utf8");
-    return { id, path: `node_modules/${pkg.name}/skills/${id}/SKILL.md`, description: content.match(/^# (.+)$/m)?.[1] || id, triggers: TRIGGERS[id] || [], rules: POLICY.baselineRules, checks: ["scan"], status: "enabled" };
+    return { id, path: `node_modules/${pkg.name}/skills/${id}/SKILL.md`, description: content.match(/^# (.+)$/m)?.[1] || id, triggers: TRIGGERS[id] || [], rules: getRules().filter((rule) => rule.skills?.includes(id)).map((rule) => rule.id), checks: ["scan"], status: "enabled" };
   });
 }
 
@@ -75,10 +75,15 @@ export function runTaskAction(action, input = {}) {
   if (action === "doctor-host") return hostDiagnostic({ ...opts, host: input.host || "unknown" });
   const profile = resolveProjectProfile({ projectRoot: opts.projectRoot, profile: input.profile }).profile;
   const routed = decideTask({ task: input.task || "", targets: opts.targets, skill: input.skill });
-  const decision = { ...routed, profile: profile.id, applicableConstraints: routed.applicable === true ? getRules({ profile: profile.id }).map((rule) => rule.id) : [] };
+  const applicable = getRules({ profile: profile.id });
+  const enabled = new Set(applicable.map((rule) => rule.id));
+  const decision = { ...routed, profile: profile.id, applicableConstraints: routed.applicable === true ? [...enabled] : [] };
+  decision.baselineRules = decision.baselineRules.filter((id) => enabled.has(id));
+  decision.requiredRules = decision.requiredRules.filter((id) => enabled.has(id));
+  decision.ruleDetails = applicable.filter((rule) => decision.requiredRules.includes(rule.id)).map((rule) => ({ id: rule.id, name: rule.title, source: `node_modules/${pkg.name}/standards/rules.json` }));
   assetReadiness(decision, opts.projectRoot);
-  if (action !== "task") return decision;
-  return { ...decision, ...core.startTask({ ...opts, task: input.task || "", decision }) };
+  if (action !== "task") return core.attachNotice(decision, opts);
+  return core.attachNotice({ ...decision, ...core.startTask({ ...opts, task: input.task || "", decision }) }, opts);
 }
 
 function hostDiagnostic(opts) {

package/bin/task-observability.cjs

@@ -265,6 +265,7 @@ function startTask(options) {
   const record = { ...baseRecord(metadata, "task-decision"), task: safeText(options.task), decision: selected, targets: strings(options.targets), inputSnapshot: captureSnapshot(options),
     ruleSnapshot: captureReadEvidence(options, [...strings(options.ruleFiles), ...strings(selected.requiredFiles)]),
     configSnapshot: contextSnapshot(options, "configFiles"), executionStatus: "not-executed", validationStatus: "unverified" };
+  attachNotice(record, options);
   record.recordPath = appendRecord(options, record);
   return record;
 }
@@ -406,12 +407,12 @@ function statusForRecords(records, options) {
   const started = records.filter((record) => record.kind === "tool-started");
   const current = validationContext(plan, tools, options);
   const failed = tools.some((record) => record.executionStatus === "failed");
-  return {
+  return attachNotice({
     recordVersion: RECORD_VERSION, packageName: options.packageName, ...lastRecordContext(records.at(-1), options),
     ...planContext(plan), records, tools,
     executionStatus: executionState(failed, started, history),
     ...current, stages: { selected: plan ? plan.decision.status : "unverified", hostDiscovered: "unverified", contentLoaded: "unverified", toolsExecuted: history.length },
-  };
+  }, options);
 }
 
 function validationContext(plan, tools, options) {
@@ -439,7 +440,7 @@ function uncoveredScope(plan, tools, options) {
 
 function planContext(plan) {
   if (!plan) return { task: null, decision: null, gaps: [] };
-  return { task: plan.task, decision: plan.decision, gaps: plan.decision.gaps || [] };
+  return { task: plan.task, decision: plan.decision, targets: plan.targets, gaps: plan.decision.gaps || [] };
 }
 
 function lastRecordContext(latest, options) {
@@ -505,18 +506,91 @@ function doctorHost(options) {
   const skills = strings(options.skillPaths).map((relative) => inspectHostFile(root, relative, Infinity));
   const gateway = options.gatewayPath ? inspectHostFile(root, options.gatewayPath, Infinity) : null;
   const warnings = entries.filter((entry) => entry.oversized).map((entry) => `${entry.path} 超过 ${maximum} 字节参考预算；请核对宿主实际配置与内容加载范围`);
-  const ready = gateway?.status === "present" && skills.length > 0 && skills.every((skill) => skill.status === "present") && entries.every((entry) => entry.status === "present" && !entry.oversized);
-  return { recordVersion: RECORD_VERSION, packageName: options.packageName, host: options.host || "unspecified", projectRoot: root, entries, skills, gateway, entryReadiness: ready ? "ready" : "incomplete", hostDiscovery: "unverified", mcpConnection: "unverified", contentLoaded: "unverified", warnings, instructionBudgetSource: "reference-default-not-observed-host-config" };
+  const runtime = inspectRuntime(root, options);
+  warnings.push(...runtime.warnings);
+  const ready = hostReady(gateway, skills, entries, runtime);
+  return { recordVersion: RECORD_VERSION, packageName: options.packageName, packageVersion: options.packageVersion, host: options.host || "unspecified", projectRoot: root, entries, skills, gateway, runtime, entryReadiness: ready ? "ready" : "incomplete", hostDiscovery: "unverified", mcpConnection: "unverified", contentLoaded: "unverified", warnings, instructionBudgetSource: "reference-default-not-observed-host-config" };
+}
+
+function hostReady(gateway, skills, entries, runtime) {
+  return gateway?.status === "present" && skills.length > 0 && skills.every((skill) => skill.status === "present") && entries.every((entry) => entry.status === "present" && !entry.oversized) && ["aligned", "unverified"].includes(runtime.status);
+}
+function runtimeVersion(root, relative, diagnostics) {
+  try { const value = JSON.parse(fs.readFileSync(readEvidencePath(root, relative), "utf8")).version; if (typeof value !== "string" || !value.trim()) throw new Error("version 必须是非空字符串"); return value; }
+  catch (error) { if (error.code !== "ENOENT") diagnostics.push(`${relative}: ${error.message}`); return null; }
+}
+function declaredDependency(root, packageName) {
+  try { const pkg = JSON.parse(fs.readFileSync(path.join(root, "package.json"), "utf8")); return pkg.dependencies?.[packageName] || pkg.devDependencies?.[packageName] || null; }
+  catch { return null; }
+}
+function runtimeMismatch(versions, declared, local) {
+  return strings(versions).length > 1 || Boolean(local && /^\d+\.\d+\.\d+(?:-[a-zA-Z0-9.-]+)?$/.test(declared || "") && declared !== local);
+}
+function inspectRuntime(root, options) {
+  root = fs.realpathSync(path.resolve(root));
+  const key = (options.packageName || "").replace("@agile-team/wl-skills-", "");
+  const manifestPath = { kit: ".wl-skills-manifest.json", design: ".wl-skills-design/state.json", test: ".wl-skills-test/manifest.json" }[key] || `.wl-skills-${key}-manifest.json`;
+  const diagnostics = [];
+  const distributedVersion = runtimeVersion(root, manifestPath, diagnostics);
+  const localVersion = runtimeVersion(root, `node_modules/${options.packageName}/package.json`, diagnostics);
+  const declaredVersion = declaredDependency(root, options.packageName);
+  const mismatch = runtimeMismatch([distributedVersion, localVersion, options.packageVersion], declaredVersion, localVersion);
+  const state = distributedVersion ? "aligned" : "unverified";
+  return { status: mismatch ? "mismatch" : diagnostics.length ? "invalid" : state, distributedVersion, localVersion, runningVersion: options.packageVersion || null, declaredVersion, diagnostics,
+    warnings: mismatch ? [`${options.packageName} 规范/本地执行器/当前执行器版本不一致：${distributedVersion} / ${localVersion} / ${options.packageVersion}；声明依赖=${declaredVersion}；同步本包依赖、锁文件与受管入口后复验`] : diagnostics, hostInvocation: "unverified" };
+}
+
+/** A notice describes the receipt; it does not claim the host displayed it or read rules. */
+function noticeProjectRoot(record, options) {
+  return options.projectRoot ? fs.realpathSync(path.resolve(options.projectRoot)) : record.notice?.projectRoot || null;
+}
+function noticeRules(item) {
+  const details = new Map((item.ruleDetails || []).map((rule) => [rule.id, rule]));
+  const ids = strings([...strings(item.baselineRules), ...strings(item.requiredRules), ...strings(item.constraints)]);
+  return ids.map((id) => ({ id, name: details.get(id)?.name || id, source: details.get(id)?.source || null }));
+}
+function noticeScope(record, options, item) {
+  return { targets: strings(options.originalTargets || record.targets || options.targets),
+    executionStatus: record.executionStatus || "not-executed", validationStatus: record.validationStatus || "unverified",
+    unverified: strings(item.unverified || ["host-discovery", "model-read-canonical-files", "planned-actions-not-executed"]) };
+}
+function noticeReason(item) { return safeText(item.reason || strings(item.reasons).join("；")); }
+function attachNotice(record, options = {}) {
+  const item = record.decision || record;
+  record.packageName ||= options.packageName;
+  record.packageVersion ||= options.packageVersion;
+  record.notice = {
+    schemaVersion: 1, packageName: record.packageName || null, packageVersion: record.packageVersion || null,
+    projectRoot: noticeProjectRoot(record, options), runId: record.runId || null, decision: item.status || "unverified",
+    skills: strings(item.selectedSkills), requiredChecks: strings(item.requiredChecks), baselineRules: strings(item.baselineRules), rules: noticeRules(item),
+    reason: noticeReason(item), missingInputs: strings(item.missingInputs),
+    gaps: (item.gaps || []).map((gap) => ({ reason: safeText(gap.reason), suggestion: safeText(gap.suggestion) })),
+    ...noticeScope(record, options, item), displayEvidence: "unverified",
+  };
+  return record;
+}
+function noticeRuleText(notice) {
+  const rules = notice.rules.slice(0, 8).map((rule) => rule.name === rule.id ? rule.id : `${rule.id}（${rule.name}）`).join("、");
+  const remaining = notice.rules.length > 8 ? `等 ${notice.rules.length} 条，完整清单见 JSON 回执` : "";
+  return `${rules}${remaining || (rules ? "" : "尚无确认规则")}`;
+}
+function noticePlanText(notice) {
+  return `${notice.requiredChecks.length ? `\n待执行检查：${notice.requiredChecks.join("、")}` : ""}` + `${notice.runId ? `\nrunId=${notice.runId}` : "\n只读判定，尚无任务记录"}；执行=${notice.executionStatus}；验证=${notice.validationStatus}`;
+}
+function noticeGapText(notice) {
+  return `${notice.missingInputs.length ? `\n缺少：${notice.missingInputs.join("、")}` : ""}` + `${notice.gaps.length ? `\n缺口：${notice.gaps.map((gap) => `${gap.reason}；建议：${gap.suggestion}`).join("\n")}` : ""}`;
 }
-
 function formatDecision(record) {
-  const item = record.decision || record;
-  const label = { baseline: "基础约束", ambiguous: "候选需确认", gap: "规则缺口", "not-applicable": "不适用", "needs-context": "缺少上下文" }[item.status] || "待判定";
-  return `[WL ${item.status}] ${strings(item.selectedSkills).join("、") || label}；${safeText(item.reason || strings(item.reasons).join("；"), 300)}${record.runId ? `；runId=${record.runId}` : ""}`;
+  const notice = record.notice || attachNotice({ ...record }).notice;
+  const label = { baseline: "基础约束", ambiguous: "候选需确认", gap: "规则缺口", "not-applicable": "不适用", "needs-context": "缺少上下文" }[notice.decision] || "待判定";
+  const identity = notice.packageName ? `${notice.packageName.replace("@agile-team/wl-skills-", "")}@${notice.packageVersion || "未知"}；` : "";
+  return `[WL ${notice.decision}] ${identity}${notice.skills.join("、") || label}；${notice.reason}` + `\n适用规则：${noticeRuleText(notice)}；范围：${notice.targets.slice(0, 3).join("、") || "未指定目标，检查范围待确认"}` + noticePlanText(notice) + noticeGapText(notice);
 }
 
 function formatStatus(status) {
-  return `[WL 执行=${status.executionStatus} 验证=${status.validationStatus}] runId=${status.runId || "尚无记录"}；工具=${status.tools.length}；待验证=${status.pendingChecks.length}；范围缺口=${(status.scopeGaps || []).length}；规则缺口=${(status.gaps || []).length}；宿主发现=${status.stages.hostDiscovered}`;
+  const identity = `${status.packageName?.replace("@agile-team/wl-skills-", "") || "未知包"}@${status.packageVersion || "未知"}`;
+  const checked = [...new Set(status.tools.flatMap((tool) => (tool.checkedFiles || []).map((file) => file.path)))];
+  return `[WL 执行=${status.executionStatus} 验证=${status.validationStatus}] ${identity}；runId=${status.runId || "尚无记录"}；工具=${status.tools.length}；实际检查文件=${checked.length}；待验证=${status.pendingChecks.length}；范围缺口=${(status.scopeGaps || []).length}；规则缺口=${(status.gaps || []).length}；证据过期=${Boolean(status.stale)}；宿主发现=${status.stages.hostDiscovered}`;
 }
 
 function aggregateStatus(statuses) {
@@ -525,4 +599,4 @@ function aggregateStatus(statuses) {
   return { recordVersion: RECORD_VERSION, runId: runIds.length === 1 ? runIds[0] : null, correlated: runIds.length === 1, packages: valid, mixedRuns: runIds.length > 1, hasFailure: valid.some((status) => status.validationStatus === "failed"), gaps: valid.flatMap((status) => status.gaps || []) };
 }
 
-module.exports = { RECORD_VERSION, STORAGE, evaluateTask, startTask, beginExecution, finishExecution, withExecution, captureSnapshot, readStatus, listGaps, doctorHost, formatDecision, formatStatus, aggregateStatus };
+module.exports = { RECORD_VERSION, STORAGE, evaluateTask, startTask, beginExecution, finishExecution, withExecution, captureSnapshot, readStatus, listGaps, doctorHost, inspectRuntime, attachNotice, formatDecision, formatStatus, aggregateStatus };

package/bin/task-observability.schema.json

@@ -34,7 +34,8 @@
     "artifacts": { "type": "array", "items": { "$ref": "#/definitions/file" } },
     "changedDuringCheck": { "type": "boolean" },
     "errorCode": { "type": "string" },
-    "recordPath": { "type": ["string", "null"] }
+    "recordPath": { "type": ["string", "null"] },
+    "notice": { "$ref": "#/definitions/notice" }
   },
   "allOf": [
     { "if": { "properties": { "kind": { "const": "task-decision" } } }, "then": { "required": ["task", "decision"], "properties": { "executionStatus": { "const": "not-executed" }, "validationStatus": { "const": "unverified" } } } },
@@ -42,6 +43,100 @@
     { "if": { "properties": { "kind": { "const": "tool-finished" } } }, "then": { "required": ["tool", "executionId", "executionStartedAt", "exitCode", "durationMs", "checks", "artifacts", "inputBefore", "changedDuringCheck"], "properties": { "executionStatus": { "enum": ["completed", "failed"] } } } }
   ],
   "definitions": {
+    "notice": {
+      "type": "object",
+      "required": [
+        "schemaVersion",
+        "packageName",
+        "packageVersion",
+        "runId",
+        "decision",
+        "skills",
+        "rules",
+        "requiredChecks",
+        "executionStatus",
+        "validationStatus",
+        "displayEvidence"
+      ],
+      "properties": {
+        "schemaVersion": {
+          "const": 1
+        },
+        "packageName": {
+          "type": [
+            "string",
+            "null"
+          ]
+        },
+        "packageVersion": {
+          "type": [
+            "string",
+            "null"
+          ]
+        },
+        "runId": {
+          "type": [
+            "string",
+            "null"
+          ]
+        },
+        "projectRoot": {
+          "type": [
+            "string",
+            "null"
+          ]
+        },
+        "decision": {
+          "type": "string"
+        },
+        "skills": {
+          "type": "array",
+          "items": {
+            "type": "string"
+          }
+        },
+        "requiredChecks": {
+          "type": "array",
+          "items": {
+            "type": "string"
+          }
+        },
+        "rules": {
+          "type": "array",
+          "items": {
+            "type": "object",
+            "required": [
+              "id",
+              "name",
+              "source"
+            ],
+            "properties": {
+              "id": {
+                "type": "string"
+              },
+              "name": {
+                "type": "string"
+              },
+              "source": {
+                "type": [
+                  "string",
+                  "null"
+                ]
+              }
+            }
+          }
+        },
+        "executionStatus": {
+          "type": "string"
+        },
+        "validationStatus": {
+          "type": "string"
+        },
+        "displayEvidence": {
+          "const": "unverified"
+        }
+      }
+    },
     "file": {
       "type": "object", "required": ["path", "sha256", "bytes"],
       "properties": { "path": { "type": "string" }, "sha256": { "type": "string", "pattern": "^[a-f0-9]{64}$" }, "bytes": { "type": "integer", "minimum": 0 } }

package/bin/wl-ui.js

@@ -574,7 +574,7 @@ function transformForEditor(content, editor) {
         ?.slice(2) || "Skill",
     );
 
-  return header + "\nStart each task with local `wl-ui task \"<task>\" --target <path>`. Read only the selected Skill. Finish actual tools with the same `--run-id` and report `wl-ui status`; model declarations do not prove verification.\n\n" + body;
+  return header + "\nStart each task with local `wl-ui task \"<task>\" --target <path>`. Before editing visibly show the returned notice: actual package/version, decision, Skill or baseline, rule IDs/names, targets, runId and unexecuted checks. Report local command failures/version drift. Read only the selected Skill. Finish actual tools with the same `--run-id` and report `wl-ui status`; model declarations do not prove verification.\n\n" + body;
 }
 
 function selectedSkills(mode, profile) {
@@ -664,7 +664,7 @@ function buildSingleFileRouter(skills, profile) {
     `> Active profile: \`${profile?.id || "native-element"}\`.`,
     "",
     "Use the package router first, then open only the relevant Skill:",
-    "Start each task with local `wl-ui task \"<task>\" --target <path>`; read only the selected canonical Skill. Finish actual checks with the same `--run-id`, then report `wl-ui status --run-id <id>`. Routing/model declarations do not prove execution or validation. Native discovery: `.agents/skills/wl-skills-ui/SKILL.md`.",
+    "Start each task with local `wl-ui task \"<task>\" --target <path>`; before editing visibly show notice (actual package/version, decision, Skill or baseline, rule IDs/names, targets, runId, unexecuted checks), including gaps and local command/version failures; read only the selected canonical Skill. Finish actual checks with the same `--run-id`, then report `wl-ui status --run-id <id>`. Routing/model declarations do not prove execution or validation. Native discovery: `.agents/skills/wl-skills-ui/SKILL.md`.",
     "Reuse one `--run-id` / `WL_TASK_RUN_ID` across installed applicable packages for this user task; do not install unused sibling packages.",
     "",
     "- `node_modules/@agile-team/wl-skills-ui/SKILL.md`",

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@agile-team/wl-skills-ui",
-  "version": "1.16.0",
+  "version": "1.17.0",
   "description": "企业级 UI 风格对齐框架 — Vue + Element Plus 双模式、任务路由与真实检查回执（tokens / element / vendors / layouts / runtime / scanner / fixer / skills）",
   "type": "module",
   "main": "./es/index.js",