analytics-dashboard
数据看板查询Skill,通过dashboard-mcp获取实时统计/系统健康/租户概览/内容队列/事件订阅等看板数据。触发:数据看板/实时统计/系统健康/租户概览/内容队列/事件订阅/早朝简报/Agent状态 不触发:内容发布/商品管理/Cookie管理
天轰穿
@thcjp
What This Skill Does
Enterprise-grade data visualization and monitoring system with custom widget builder, advanced analytics (trend, anomaly, prediction), multi-channel alerts, report export, team collaboration, custom data sources, theme customization, and real-time data push.
Replaces manually building dashboards from scattered data sources by providing a unified, team-ready platform with automated alerts and scheduled report generation.
When to Use It
- Create a real-time operations dashboard combining email trends, task status, and SLA metrics
- Set up multi-channel alerts for anomaly detection on key business metrics
- Schedule daily PDF or Excel report exports and email them to stakeholders
- Build custom widgets that aggregate data from multiple sources with conditional formatting
- Share interactive dashboards with team members using role-based access control
- Monitor live data streams via WebSocket push without manual page refreshes
Install
$ openclaw skills install @thcjp/analytics-dashboard核心功能: 本技能提供/Cookie管理等能力。
Analytics Dashboard Skill
数据看板查询引擎,通过dashboard-mcp统一获取系统运营数据看板(R-82 Skill连接补全)。状态: 🟢已实现 | 优先级: P1
使用场景
- 实时看板: 查询系统实时统计(活跃状态/内容管道数/告警数)
- 系统健康: 检查Docker/Gateway/磁盘健康状态
- 租户概览: 查询指定租户的发布量和最近发布记录
- 内容队列: 查询租户内容队列状态(待处理/处理中/完成/失败)
- 事件订阅: 获取最近N条系统事件(告警/KPI/设备)
- Agent状态: 查询所有Agent运行状态
- 早朝简报: 获取每日早朝简报数据
工作流
实时看板查询(realtime)
- 接收请求(action=realtime)
- 调用dashboard-mcp的
get_realtime_stats工具获取实时统计 - 返回live_status + content_pipelines_24h + pending_alerts
系统健康检查(health)
- 接收请求(action=health)
- 调用dashboard-mcp的
get_system_health工具获取健康状态 - 返回Docker/Gateway/Disk健康状态+overall判定(healthy/degraded)
租户概览查询(tenant)
- 接收请求(action=tenant, tenant_id=<参数>)
- 调用dashboard-mcp的
get_tenant_overview工具,传入tenant_id - 返回租户信息+总发布量+最近5条发布记录
内容队列状态(queue)
- 接收请求(action=queue, tenant_id=<参数>)
- 调用dashboard-mcp的
get_content_queue_status工具,传入tenant_id - 返回队列总数+各状态计数(pending/processing/completed/failed)
事件订阅(events)
- 接收请求(action=events, event_type=all, limit=50)
- 调用dashboard-mcp的
subscribe_events工具,传入event_type和limit - 返回最近N条事件(最新在前)
Agent状态查询(agents)
- 接收请求(action=agents)
- 调用dashboard-mcp的
get_agents_status工具获取所有Agent状态 - 返回Agent总数+各Agent状态+live_status
数据源
统一数据查询入口:通过dashboard-mcp(R-82)获取看板数据,禁止Skill自行读取data/下数据文件,消除碎片化
| 数据类型 | 获取方式 | dashboard-mcp工具 |
|---|---|---|
| 实时统计 | dashboard-mcp get_realtime_stats | 聚合live_status+管道数+告警数 |
| 系统健康 | dashboard-mcp get_system_health | Docker/Gateway/Disk(复用health_checks) |
| 租户概览 | dashboard-mcp get_tenant_overview | 发布量+最近发布 |
| 内容队列 | dashboard-mcp get_content_queue_status | 队列状态计数 |
| 事件订阅 | dashboard-mcp subscribe_events | 最近N条事件(alert_queue.jsonl) |
| Agent状态 | dashboard-mcp get_agents_status | Agent列表+live_status |
异常处理
| 异常 | 错误码 | 处理 |
|---|---|---|
| dashboard-mcp不可用 | MCP_UNAVAILABLE | 返回降级提示,建议检查dashboard-mcp容器状态 |
| tenant_id为空 | TENANT_ID_EMPTY | 提示"租户ID必填" |
| 数据文件不存在 | DATA_NOT_FOUND | 返回空数据+提示"暂无数据" |
| MCP调用超时 | MCP_TIMEOUT | circuit_breaker熔断,返回最近缓存数据 |
| 工具参数无效 | INVALID_PARAMS | 提示参数格式要求 |
输入格式
{
"action": "realtime|health|tenant|queue|events|agents",
"tenant_id": "<参数>",
"event_type": "all|kpi|alert|device",
"limit": 50
}
字段说明:
action: 操作类型(realtime实时看板/health系统健康/tenant租户概览/queue内容队列/events事件订阅/agents状态查询)tenant_id: 租户ID(tenant/queue操作必填)event_type: 事件类型(events操作使用,默认all)limit: 返回条数(events操作使用,默认50,最大500)
输出格式
{
"success": true,
"data": {
"action": "realtime",
"live_status": {},
"content_pipelines_24h": 5,
"pending_alerts": 2,
"timestamp": "2026-07-17T10:00:00Z"
},
"error": null,
"code": null
}
字段说明:
live_status: 系统活跃状态对象content_pipelines_24h: 近24小时内容管道数pending_alerts: 待处理告警数timestamp: 数据获取时间(ISO格式)
示例
示例1: 实时看板查询
- 调用realtime → dashboard-mcp
get_realtime_stats→ 返回实时统计 - 返回:
{success:true, data:{live_status:{}, content_pipelines_24h:5, pending_alerts:2, timestamp:"..."}}
示例2: 租户概览查询
- 调用tenant(tenant_id="t001") → dashboard-mcp
get_tenant_overview("t001")→ 返回租户概览 - 返回:
{success:true, data:{tenant_id:"t001", total_publishes:42, recent_publishes:[...], timestamp:"..."}}
示例3: 系统健康检查
- 调用health → dashboard-mcp
get_system_health→ 返回健康状态 - 返回:
{success:true, data:{docker:{healthy:true}, gateway:{healthy:true}, disk:{}, overall:"healthy", timestamp:"..."}}
关联文档
- dashboard-mcp: mcps/dashboard-mcp/server.py (20个工具, DEF-85阶段2)
- 05文档§四 DEF-07 (dashboard迁移设计)
- SKILL_REGISTRY: agents/hubu/SKILL_REGISTRY.md (已注册)
变更历史
| 版本 | 日期 | 变更说明 |
|---|---|---|
| v1.0.0 | 2026-07-17 | B2-09: 创建analytics-dashboard Skill连接dashboard-mcp(R-82), 封装6核心看板查询工具, 补全hubu SKILL_REGISTRY已注册但缺失的Skill文件 |
Top skills in this category
API Gateway
@byungkyuCall third-party APIs through the Maton gateway, which injects the credential for an app the user has already connected. Use this skill when the user names a connected app and a concrete action in it - read a mailbox, query a CRM, file an issue, update a spreadsheet, etc.
Marketing Skills
@jchopard69Access 23 marketing modules offering checklists, frameworks, and ready-to-use deliverables for CRO, SEO, copywriting, analytics, launches, ads, and social me...
OpenClaw Command Center
@jontsaiMission control dashboard for OpenClaw - real-time session monitoring, LLM usage tracking, cost intelligence, and system vitals. View all your AI agents in one place.
百度网盘
@wscats百度网盘(Baidu Drive, pan.baidu.com)文件管理 — 上传、下载、转存、分享、搜索、移动、复制、重命名、创建文件夹、删除(高风险,需用户确认)。 同时支持 Agent 记忆备份/恢复(kimiclaw/maxclaw/qclaw/openclaw)。 TRIGGER: 用户消息明确提及"百...
moltbook-interact
@lunarcmdInteract with Moltbook social network for AI agents. Post, reply, browse, and analyze engagement. Use when the user wants to engage with Moltbook, check their feed, reply to posts, or track their activity on the agent social network.