performance-monitor

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Performance Monitor

绩效监控

One monitor skill, two modes. report
builds a stakeholder-facing multi-metric snapshot of what already happened (traffic, rankings, GEO/AI, authority, backlinks, content) and turns period-over-period deltas into prioritized recommendations. alert
configures forward-looking threshold and anomaly notifications so a drop in rankings, traffic, technical health, backlinks, competitor position, or AI citations fires before someone eyeballs it. Pick the mode from intent: past-tense reporting →

report

; future-tense "tell me when" →

alert

Mode set:

report

(multi-metric snapshot) ·

alert

(forward thresholds/anomalies). Default when unstated: infer from verb tense (see Decision Gates).

Scope guard — what this skill does NOT do: it does not compute the CORE-EEAT content score or run its vetoes (T04/C01/R10) — that gate is content-quality-auditor; it does not compute the CITE domain score or run its vetoes (T03/T05/T09) — that gate is domain-authority-auditor. This skill reports those scores when a gate has already produced them and watches them for change; it never scores. Raw position-by-position ranking deltas belong to rank-tracker.

这是一个具备两种模式的监控Skill。**

report

模式可面向利益相关方构建已发生情况的多指标快照（涵盖流量、排名、地域/AI表现、权重、反向链接、内容），并将周期间的差值转化为优先级明确的建议。

alert

**模式可配置前瞻性的阈值和异常通知，这样当排名、流量、技术健康度、反向链接、竞品排名或AI引用出现下降时，能在人工发现前触发警报。根据意图选择模式：过去时态的报告需求→

report

；未来时态的「当X发生时通知我」需求→

alert

。

模式设置：

report

（多指标快照）·

alert

（前瞻性阈值/异常）。未明确指定时的默认规则：根据动词时态推断（参见决策 gate）。

范围限制——本Skill不负责的内容：不计算CORE-EEAT内容评分或执行其否决规则（T04/C01/R10）——该功能由content-quality-auditor负责；不计算CITE域名评分或执行其否决规则（T03/T05/T09）——该功能由domain-authority-auditor负责。本Skill仅在相关gate已生成上述评分时进行报告，并监控其变化；从不自行计算评分。逐位的原始排名差值属于rank-tracker的功能范围。

Quick Start

快速开始

text

undefined

text

undefined

report mode

report模式

Create an SEO performance report for [domain] for [time period] Generate an executive summary of SEO performance for [month/quarter] Create a GEO visibility report for [domain]

为[域名]创建[时间段]的SEO绩效报告生成[月度/季度]SEO绩效的执行摘要为[域名]创建地域可见性报告

alert mode

alert模式

Set up SEO monitoring alerts for [domain] Create ranking drop alerts for my top 20 keywords Alert me when AI citations for [domain] drop


Shortest valid invocation: `performance-monitor <domain>` (mode inferred). Output: **report** returns a metric-table → what-changed → why → next-action dashboard with every figure source-tagged; **alert** returns an alert-config summary with named triggers, thresholds, priorities, and delivery routing. Both emit a handoff summary ready for `memory/monitoring/`.

为[域名]设置SEO监控警报为我的前20个关键词创建排名下降警报当[域名]的AI引用下降时通知我


最短有效调用：`performance-monitor <域名>`（模式自动推断）。输出：**report**模式返回包含指标表格→变化内容→原因→下一步行动的仪表盘，每个数据都标记来源；**alert**模式返回包含命名触发器、阈值、优先级和交付路径的警报配置摘要。两种模式都会生成可提交至`memory/monitoring/`的交接摘要。

Skill Contract

Skill协议

Expected output: mode

report

→ a delta-based multi-metric report/dashboard; mode

alert

→ an alert configuration summary. Both include the standard handoff summary for

memory/monitoring/

Reads: prior baselines and current performance data.
```
report
```
reads current + prior-period metrics across traffic/rankings/authority/content, report audience, and date range.
```
alert
```
reads baselines, critical keywords/metrics to watch, normal volatility, and delivery preferences. Plus any user-provided or tool data.
Writes: a user-facing monitoring deliverable plus a reusable summary storable under
```
memory/monitoring/
```
.
Promotes: significant changes, confirmed anomalies, durable thresholds, and follow-up actions; pending decisions go to
```
memory/open-loops.md
```
(never directly to
```
decisions.md
```
).
Done when:
- ```
report
```
  : each in-scope section (traffic, rankings, GEO, authority, backlinks, content) is present or marked "Not yet evaluated"; every metric is source-tagged and compared to the prior period; recommendations carry owner, priority, and expected impact.
- ```
alert
```
  : each chosen alert category has a named trigger, threshold, and priority; a Critical/High/Medium/Low response plan and delivery routing are defined; thresholds are tuned to the metric's stated normal volatility.
Primary next skill: see Next Best Skill — the two modes hand off to each other.

预期输出：

report

模式→基于差值的多指标报告/仪表盘；

alert

模式→警报配置摘要。两者均包含适用于

memory/monitoring/

的标准交接摘要。

读取：历史基线和当前绩效数据。
```
report
```
模式读取流量/排名/权重/内容的当前及历史周期指标、报告受众和日期范围。
```
alert
```
模式读取基线数据、需监控的关键关键词/指标、正常波动范围和交付偏好。此外还包括用户提供的任何数据或工具数据。
写入：面向用户的监控交付成果，以及可存储在
```
memory/monitoring/
```
下的可复用摘要。
推送：重大变化、已确认的异常、固定阈值和后续行动；待决策项存入
```
memory/open-loops.md
```
（绝不直接存入
```
decisions.md
```
）。
完成条件：
- ```
report
```
  模式：所有范围内的板块（流量、排名、地域、权重、反向链接、内容）均已呈现或标记为「尚未评估」；每个指标都标记来源并与历史周期对比；建议包含负责人、优先级和预期影响。
- ```
alert
```
  模式：每个选定的警报类别都有命名触发器、阈值和优先级；已定义紧急/高/中/低响应计划和交付路径；阈值已根据指标的既定正常波动范围调整。
主要后续Skill：参见最佳后续Skill——两种模式可互相交接。

Handoff Summary

交接摘要

Emit the standard shape from skill-contract.md §Handoff Summary Format. Name the active mode (
report
/
alert
) in the Objective line.

按照skill-contract.md §交接摘要格式生成标准格式。在目标行中注明当前激活的模式（
report
/
alert
）。

Data Sources

数据源

All integrations optional (see CONNECTORS.md). Tier 1 (keyless) works for both modes; keyed tools are opt-in Tier 2/3.

report: with tools connected, aggregates traffic from ~~analytics, search data from ~~search console, rankings/backlinks from ~~SEO tool, and AI visibility from ~~AI monitor. Without tools, ask the user for analytics exports, Search Console data, ranking data, and KPIs.
alert: with tools, monitor real-time feeds from ~~SEO tool, ~~search console, and ~~web crawler. Without tools, ask for baselines, critical keywords, delivery preferences, and historical data.

Zero-dependency measurement loop (both modes): every reported change or fired alert should come from a computed delta, not an eyeballed estimate. Store each period's KPIs and let the ledger compute movement:

python3 "${CLAUDE_PLUGIN_ROOT}/scripts/connectors/ledger.py" record <domain> --source monitor --data '{"sessions": ..., "clicks": ..., ...}'

, then

ledger.py diff <domain> --source monitor

for the period delta and

ledger.py trend <domain> --source monitor --field <kpi>

for the trend line. Label every figure Measured / User-provided / Estimated, and attribute outcome movement against a control rather than to the latest change — see references/measurement-protocol.md. See scripts/connectors/README.md.

所有集成均为可选（参见CONNECTORS.md）。Tier 1（无密钥）适用于两种模式；需密钥的工具为可选的Tier 2/3。

report模式：连接工具后，可从~~analytics聚合流量数据、从~~search console获取搜索数据、从~~SEO tool获取排名/反向链接数据、从~~AI monitor获取AI可见性数据。未连接工具时，需向用户请求分析导出文件、Search Console数据、排名数据和关键绩效指标（KPI）。
alert模式：连接工具后，可监控来自~~SEO tool、~~search console和~~web crawler的实时数据馈送。未连接工具时，需向用户请求基线数据、关键关键词、交付偏好和历史数据。

零依赖测量循环（两种模式适用）：所有报告的变化或触发的警报都应来自计算得出的差值，而非人工估算。存储每个周期的KPI，由分类账计算变化：

python3 "${CLAUDE_PLUGIN_ROOT}/scripts/connectors/ledger.py" record <域名> --source monitor --data '{"sessions": ..., "clicks": ..., ...}'

，然后使用

ledger.py diff <域名> --source monitor

获取周期差值，使用

ledger.py trend <域名> --source monitor --field <kpi>

获取趋势线。为每个数据标记「测量所得」「用户提供」「估算」，并对照对照组而非最新变化来归因结果变动——参见references/measurement-protocol.md。详情请见scripts/connectors/README.md。

Decision Gates

决策Gate

Stop and ask the user when:

Mode ambiguous — the request fits neither past-tense ("report on last month") nor future-tense ("tell me when X drops"). Offer: (1)
```
report
```
a snapshot now, (2)
```
alert
```
for ongoing monitoring, (3) both.
report: no reporting period or comparison period can be determined and none is in context — offer (1) last 30 days vs prior 30, (2) last calendar month vs prior month, (3) a custom range. A period comparison is required, not optional.
alert: no baseline or normal-volatility reference exists for the metrics to watch — thresholds would be arbitrary. Offer (1) supply recent baseline data, (2) use the Alert Threshold Quick Reference defaults and label them Estimated, (3) cancel.

Continue silently (never stop for):

A report section's source data is missing — mark that section "Not yet evaluated" and proceed; do not fabricate the metric.
Audience not stated (report) — default to the executive template and note the assumption.
Delivery channel not stated (alert) — default to the channel in context (or email) and note it.
An alert category the user did not mention — leave it unconfigured; do not add alerts they did not request.

需停止并询问用户的情况：

模式模糊——请求既不符合过去时态（「报告上月情况」）也不符合未来时态（「当X下降时通知我」）。提供选项：(1) 立即生成
```
report
```
快照，(2) 设置
```
alert
```
进行持续监控，(3) 两者都选。
report模式：无法确定报告周期或对比周期，且上下文无相关信息——提供选项：(1) 过去30天对比之前30天，(2) 上一自然月对比前一个月，(3) 自定义范围。必须进行周期对比，不可省略。
alert模式：需监控的指标无基线或正常波动参考——阈值将是任意的。提供选项：(1) 提供近期基线数据，(2) 使用警报阈值快速参考的默认值并标记为「估算」，(3) 取消操作。

可静默继续（无需停止）的情况：

报告板块的源数据缺失——将该板块标记为「尚未评估」并继续；不得编造指标。
未指定受众（report模式）——默认使用高管模板并注明该假设。
未指定交付渠道（alert模式）——默认使用上下文指定的渠道（或邮件）并注明。
用户未提及的警报类别——保持未配置状态；不得添加用户未请求的警报。

Instructions

操作说明

Step 0 — Select mode. Read

--mode

if given. Otherwise infer: past-tense / "how did we do" / "月报" →

report

; future-tense / "tell me when" / "预警" →

alert

. If neither fits, use the mode-ambiguous gate above.

步骤0——选择模式：若指定了

--mode

则读取该参数。否则自动推断：过去时态/「我们表现如何」/「月报」→

report

；未来时态/「当X发生时通知我」/「预警」→

alert

。若两者都不符合，使用上述模式模糊的处理规则。

Mode: report

模式：report

Use Report Output Templates and cover:

Define Report Parameters — domain, period, comparison period, report type, audience, focus areas, data freshness.
Executive Summary — overall rating, wins, watch areas, required actions, metrics-at-a-glance (traffic, rankings, conversions, DA/authority, AI citations), and SEO ROI; tag each metric Measured / User-provided / Estimated.
Organic Traffic — sessions, users, pageviews, engagement/bounce, trend, source/device split, top pages, each figure source-tagged.
Keyword Rankings — position ranges, distribution change, top improvements/declines, SERP features. For raw position-by-position deltas, defer to rank-tracker rather than recomputing here.
GEO / AI Performance — AI citation overview, citations by topic, GEO wins, optimization opportunities.
Domain Authority (CITE) — include CITE dimension scores and veto status when a gate has already produced them; otherwise mark "Not yet evaluated." Do not compute CITE here.
Content Quality (CORE-EEAT) — include average scores and trends when already produced; otherwise mark "Not yet evaluated." Do not compute CORE-EEAT here.
Backlinks — link-profile summary, acquisition trend, notable links, competitive position.
Content Performance — publishing summary, top content, content needing attention, content ROI.
Recommendations — immediate/short-term/long-term actions with priority, owner, expected impact, next-period goals.
Compile Full Report — table of contents, appendix, data sources, methodology, glossary.

使用报告输出模板，涵盖以下内容：

定义报告参数——域名、周期、对比周期、报告类型、受众、重点领域、数据新鲜度。
执行摘要——总体评级、亮点、需关注领域、必要行动、关键指标概览（流量、排名、转化、域名权重/权威度、AI引用）以及SEO投资回报率（ROI）；为每个指标标记「测量所得」「用户提供」「估算」。
自然流量——会话数、用户数、页面浏览量、参与度/跳出率、趋势、来源/设备分布、热门页面，每个数据都标记来源。
关键词排名——排名区间、分布变化、排名提升/下降最多的关键词、搜索结果页（SERP）功能。逐位的原始排名差值请交由rank-tracker处理，请勿在此重新计算。
地域/AI表现——AI引用概述、按主题划分的引用、地域表现亮点、优化机会。
域名权威度（CITE）——若相关gate已生成CITE维度评分和否决状态，则包含该内容；否则标记为「尚未评估」。请勿在此计算CITE评分。
内容质量（CORE-EEAT）——若已生成平均评分和趋势，则包含该内容；否则标记为「尚未评估」。请勿在此计算CORE-EEAT评分。
反向链接——链接配置文件摘要、获取趋势、重要链接、竞品位置。
内容绩效——发布摘要、热门内容、需关注的内容、内容ROI。
建议——即时/短期/长期行动，包含优先级、负责人、预期影响、下一周期目标。
编译完整报告——目录、附录、数据源、方法论、术语表。

Mode: alert

模式：alert

Use Alert Configuration Templates and:

Define Alert Categories — choose from rankings, traffic, technical, backlinks, competitors, GEO / AI, and brand.
Configure Alert Rules by Category — define trigger condition, threshold, alert name, and priority for each relevant rule; tie each threshold to a stated baseline and label that baseline Measured / User-provided / Estimated.
Define Alert Response Plans — map Critical / High / Medium / Low to response time and next actions.
Set Up Alert Delivery — channels, routing, cooldowns, maintenance windows, escalation paths.
Create Alert Summary — output category counts, the critical playbook, and a weekly review checklist as the deliverable.

使用警报配置模板，完成以下操作：

定义警报类别——从排名、流量、技术、反向链接、竞品、地域/AI、品牌中选择。
按类别配置警报规则——为每个相关规则定义触发条件、阈值、警报名称和优先级；将每个阈值与既定基线关联，并标记该基线为「测量所得」「用户提供」「估算」。
定义警报响应计划——将紧急/高/中/低优先级映射到响应时间和下一步行动。
设置警报交付——渠道、路由、冷却时间、维护窗口、升级路径。
创建警报摘要——输出类别统计、紧急应对手册和每周检查清单作为交付成果。

Shared discipline (both modes)

通用规则（两种模式）

Label every metric Measured (tool/export), User-provided, or Estimated (model inference); never present an estimate as measured; if a required metric is unavailable, mark it N/A — do not invent it. Separate an observed change from a plausible explanation (corroborate before stating a cause), an optimization opportunity, and follow-up needing crawl/SERP/rank/audit — never report an unverified explanation as a confirmed cause.

为每个指标标记测量所得（工具/导出）、用户提供或估算（模型推断）；绝不能将估算值作为测量值呈现；若所需指标不可用，标记为N/A——不得编造指标。将观察到的变化与合理的解释（陈述原因前需证实）、优化机会和需爬取/搜索结果页/排名/审计的后续行动分开；绝不能将未验证的解释作为已确认的原因报告。

Alert Threshold Quick Reference

警报阈值快速参考

Metric	Warning	Critical	Frequency
Organic traffic	-15% WoW	-30% WoW	Daily
Keyword positions	>3 position drop	>5 position drop	Daily
Pages indexed	-5% change	-20% change	Weekly
Crawl errors	>10 new/day	>50 new/day	Daily
Core Web Vitals	"Needs Improvement"	"Poor"	Weekly
Backlinks lost	>5% in 1 week	>15% in 1 week	Weekly
AI citation loss	Any key query	>20% queries	Weekly
Security issues	Any detected	Any detected	Daily

Steep-decline trigger (always on): if organic traffic OR aggregate keyword rank falls >30% below its trailing baseline (default: prior 28-day median for the same weekday band), fire a Critical alert regardless of category. Use a trailing median, not a single prior day, so one noisy data point does not trip it. Label the baseline Measured / User-provided / Estimated.

指标	警告	紧急	频率
自然流量	周环比下降15%	周环比下降30%	每日
关键词排名	下降超过3位	下降超过5位	每日
已索引页面	变化率-5%	变化率-20%	每周
爬取错误	每日新增超过10个	每日新增超过50个	每日
Core Web Vitals	"需要改进"	"较差"	每周
丢失的反向链接	1周内丢失超过5%	1周内丢失超过15%	每周
AI引用丢失	任意关键查询	超过20%的查询	每周
安全问题	检测到任何问题	检测到任何问题	每日

陡降触发器（始终启用）：若自然流量或综合关键词排名低于其历史基线（默认：过去28天同一工作日时段的中位数）超过30%，无论类别如何，均触发紧急警报。使用历史中位数而非单一历史数据，避免单个异常数据点触发警报。标记基线为「测量所得」「用户提供」「估算」。

Reading Deltas Against a Control

对照对照组分析差值

A reported delta or a fired alert is only evidence if it beats a control over a fixed readback window set before the change — a raw before/after on a confounded outcome is a story, not proof. Attach the decision protocol from references/measurement-protocol.md §Cross-discipline decision protocol:

Readback window — pick the window for the change type up front (content refresh 7/14/28/56 days; new content 14/28/56/90; technical fix daily ×7 then 28; AEO/GEO surfacing weekly) and do not react to noise inside it. A fired alert opens a readback window, not an instant verdict — confirm the drop holds before declaring an incident.
Required readback fields — record: change · owner · baseline window · candidate window · sources · primary + secondary metric · winner · caveats · decision · next-patch · next-readback date.
Decision — mark each change Promote (beats control on the primary metric past the bar), Keep-testing (trending, not yet significant), Rollback (loses by the same bar), or Unproven (everything else). Report delta-vs-control, not raw delta.

仅当报告的差值或触发的警报在变更前设定的固定回读窗口内优于对照组时，才视为有效证据——受干扰结果的原始前后对比只是故事，而非证明。附上references/measurement-protocol.md §跨领域决策协议中的决策流程：

回读窗口——提前为变更类型选择窗口（内容更新：7/14/28/56天；新内容：14/28/56/90天；技术修复：连续7天每日监测，之后28天监测；AEO/地域展示：每周），窗口内不对噪音做出反应。触发警报意味着开启回读窗口，而非立即判定——确认下降趋势持续后再宣布为事件。
必填回读字段——记录：变更·负责人·基线窗口·候选窗口·来源·主要+次要指标·结果·注意事项·决策·下次补丁·下次回读日期。
决策——将每个变更标记为推广（主要指标优于对照组且达标）、持续测试（有趋势但未达显著水平）、回滚（未达标且表现更差）或未验证（其他情况）。报告与对照组的差值，而非原始差值。

Example

示例

report: an executive summary with overall status, metrics-at-a-glance for traffic/rankings/conversions/authority/AI citations, SEO ROI, and immediate/month/quarter actions with owners and dates.
alert: a keyword alert matrix with Critical vs High thresholds, a response plan for drops, and notification routing to email + Slack.

report模式：包含总体状态、流量/排名/转化/权威度/AI引用的关键指标概览、SEO ROI，以及带负责人和日期的即时/月度/季度行动的执行摘要。
alert模式：包含紧急/高阈值的关键词警报矩阵、排名下降的响应计划，以及邮件+Slack的通知路由。

Save Results

保存结果

Ask "Save these results?" If yes, write to

memory/monitoring/

using filename

YYYY-MM-DD-<topic>.md

— see skill-contract.md §Save Results Template. This is a non-auditor skill: ask before writing memory and hand off veto-like risks to the relevant auditor gate rather than appending veto markers itself.

询问用户「是否保存这些结果？」。若同意，使用文件名

YYYY-MM-DD-<主题>.md

写入

memory/monitoring/

——参见skill-contract.md §保存结果模板。本Skill非审计类Skill：写入内存前需询问用户，并将类似否决的风险移交至相关审计gate，而非自行添加否决标记。

Reference Materials

参考资料

Report Output Templates — compact starter blocks for all 11 report sections (report mode)
KPI Definitions — SEO/GEO metric definitions with benchmarks, thresholds, trend analysis, attribution guidance
Report Templates by Audience — copy-ready templates for executive, marketing, technical, and client audiences
Alert Configuration Templates — full category tables, thresholds, response-plan templates (alert mode)
Alert Threshold Guide — threshold setting, fatigue prevention, escalation paths, response playbooks
Measurement & Attribution Protocol — readback windows, required fields, and the promote / keep-testing / rollback / unproven decision rule

报告输出模板——所有11个报告板块的精简起始模块（report模式）
KPI定义——SEO/地域指标的定义、基准、阈值、趋势分析、归因指南
按受众划分的报告模板——适用于高管、营销、技术和客户受众的可直接使用的模板
警报配置模板——完整的类别表格、阈值、响应计划模板（alert模式）
警报阈值指南——阈值设置、防止警报疲劳、升级路径、响应手册
测量与归因协议——回读窗口、必填字段，以及推广/持续测试/回滚/未验证的决策规则

Next Best Skill

最佳后续Skill

Mode-conditional, then terminal:

After report — a change needs ongoing monitoring → run this skill in
```
alert
```
mode. A section marked "Not yet evaluated" for authority → domain-authority-auditor; for content quality → content-quality-auditor. One-off report with no action → Terminal.
After alert — a reporting cadence is requested → run this skill in
```
report
```
mode. Standalone alert setup → Terminal.

Termination: the visited-set and

max-depth: 3

rules from skill-contract.md §Termination rules apply. Do not re-enter a mode already run in this chain (report→alert→report is a visited-set stop); if routing is ambiguous, present the options and stop instead of auto-following.

取决于当前模式，之后终止流程：

report模式之后——某变化需要持续监控→以
```
alert
```
模式运行本Skill。权威度板块标记为「尚未评估」→domain-authority-auditor；内容质量板块标记为「尚未评估」→content-quality-auditor。一次性报告且无后续行动→终止。
alert模式之后——要求定期报告→以
```
report
```
模式运行本Skill。独立警报设置完成→终止。

终止规则：遵循skill-contract.md §终止规则中的已访问集合和

max-depth: 3

规则。不得重新进入当前流程中已运行过的模式（report→alert→report属于已访问集合，需停止）；若路由模糊，呈现选项后停止，而非自动跳转。