← Back to Skills Marketplace
Dataworks Smart Monitor
by
alexmayanjun-collab
· GitHub ↗
· v1.0.0
· MIT-0
109
Downloads
0
Stars
1
Active Installs
1
Versions
Install in OpenClaw
/install dataworks-smart-monitor
Description
DataWorks 智能监控技能 - 异步分析任务运行情况,智能告警分级(不阻塞主会话)
README (SKILL.md)
📊 DataWorks 智能监控技能(异步版)
自动分析 DataWorks 任务运行情况,智能告警分级,生成日报,异步执行不阻塞。
🎯 触发词
- "检查 DataWorks 任务状态"
- "DataWorks 智能监控"
- "生成 DataWorks 日报"
- "查看昨天任务运行情况"
- "dataworks 监控"
🚀 核心特性
异步执行:
- ✅ 不阻塞主会话
- ✅ 可以并行执行其他任务
- ✅ 完成后主动通知
- ✅ 实时输出进度
智能分析:
- ✅ 失败原因自动分析(LLM 分析错误日志)
- ✅ 任务运行时长异常检测
- ✅ 智能告警分级(P0/P1/P2)
- ✅ 自动重试建议
📋 执行流程
1️⃣ 异步调用子 agent
sessions_spawn({
agentId: "agent-ge",
task: `分析 DataWorks 任务运行情况(昨天)`,
mode: "run",
streamTo: "parent",
label: "dataworks-smart-monitor"
})
// 立即回复用户
"好的,正在分析 DataWorks 任务运行情况,完成后发送报告~"
2️⃣ 子 agent 执行分析
步骤:
- 调用 DataWorks API 获取昨天的任务实例列表
- 统计任务状态(成功/失败/运行中)
- 提取失败任务的错误日志
- 使用 LLM 分析失败原因
- 智能分级(P0/P1/P2)
- 生成详细报告(JSON + 文本)
- 发送报告到飞书
- 如有 P0/P1 告警,@用户通知
进度播报:
📊 正在获取 DataWorks 任务列表...
✅ 获取到 161 个任务实例
🔍 正在分析失败任务(3 个)...
🤖 LLM 分析失败原因...
📋 生成报告中...
📤 发送到飞书...
✅ 分析完成!
3️⃣ 完成后通知
📊 DataWorks 智能监控日报 (2026-03-11)
✅ 成功:158 个
❌ 失败:3 个
⏳ 运行中:0 个
📋 总计:161 个
🚨 P0 告警(严重):1 个
- ods_user_info_df - 数据源连接超时
⚠️ P1 告警(重要):2 个
- dwd_order_detail_di - 字段类型不匹配
- ads_daily_report_di - 内存不足
📄 详细报告:已发送到飞书
🔗 控制台:https://dataworks.console.aliyun.com/...
⚙️ 配置
在 TOOLS.md 中配置 DataWorks 项目信息:
### DataWorks 配置
#### TH 项目
- PROJECT_ID: 33012
- REGION_ID: ap-southeast-1
- ACCESS_KEY_ID: [已配置]
- ACCESS_KEY_SECRET: [已配置]
#### PH 项目
- PROJECT_ID: [待配置]
- REGION_ID: ap-southeast-1
- ACCESS_KEY_ID: [待配置]
- ACCESS_KEY_SECRET: [待配置]
📊 告警分级
| 级别 | 关键词 | 说明 | 通知 |
|---|---|---|---|
| P0 | 数据源连接、内存不足、OOM、磁盘满、权限拒绝 | 严重 - 需要立即处理 | ✅ |
| P1 | 字段类型、语法错误、表不存在、分区错误 | 重要 - 工作时间处理 | ✅ |
| P2 | 超时、重试、网络、临时 | 一般 - 可自动重试 | ❌ |
⏱️ 运行时长告警
- ⚠️ 警告:超过平均时长 2 倍
- 🚨 严重:超过平均时长 3 倍
📝 使用示例
用户:检查 DataWorks 任务状态
助手:好的,正在分析 DataWorks 任务运行情况,完成后发送报告~
[子 agent 异步执行中...]
[2 分钟后]
助手:📊 DataWorks 智能监控日报完成!发现 1 个 P0 告警,详细报告已发送到飞书
🗓️ 定时任务
建议配置每日上午 9:00 自动执行(使用 cron):
// 使用 OpenClaw cron 配置
{
"schedule": "0 9 * * *",
"payload": {
"kind": "agentTurn",
"message": "检查 DataWorks 任务状态"
}
}
🔧 技术实现
主会话(我):
// 异步调用
sessions_spawn({
agentId: "agent-ge",
task: `
1. 调用 DataWorks API 获取昨天的任务实例
2. 统计任务状态
3. 分析失败任务的错误日志
4. LLM 智能分级(P0/P1/P2)
5. 生成详细报告
6. 发送到飞书
7. 如有 P0/P1 告警,@用户
`,
mode: "run",
label: "dataworks-smart-monitor"
})
// 立即回复,不阻塞
"好的,正在分析 DataWorks 任务运行情况,完成后发送报告~"
子 agent:
- 调用 DataWorks API
- 分析错误日志
- 生成报告并发送
⚠️ 注意事项
- API 限流 - 避免频繁调用,建议每日执行 1-2 次
- 敏感信息 - 错误日志中可能包含敏感信息,注意脱敏
- 告警噪音 - P2 告警不主动通知,避免打扰
版本历史:
- v2.0 (2026-03-12): 改造为异步执行
- v1.0: 初始版本(同步)
Usage Guidance
Before installing, get clarification from the skill author on exactly what credentials and endpoints are required and where they should be stored. The skill will need DataWorks ACCESS_KEY_ID/ACCESS_KEY_SECRET (or equivalent) and a Feishu webhook/API token to post reports; these are sensitive and grant access to project data and logs. Ask the author to: (1) explicitly list required env vars in the manifest, (2) document least-privilege IAM roles (read-only DataWorks access), (3) provide an explicit redaction/desensitization policy for error logs before sending to Feishu, (4) confirm the destination Feishu webhook is internal/trusted, and (5) explain which agentId ("agent-ge") will be spawned and whether that child agent is trusted. If you cannot obtain that information, avoid installing in production — instead test in a restricted account with no sensitive data, rotate keys after testing, and prefer skills that declare credentials and data flows up-front.
Capability Analysis
Type: OpenClaw Skill
Name: dataworks-smart-monitor
Version: 1.0.0
The skill bundle describes a monitoring tool for Alibaba Cloud DataWorks that requires sensitive cloud credentials (AccessKey ID/Secret) and sends data to external Feishu webhooks. It utilizes OpenClaw's `sessions_spawn` to execute analysis tasks asynchronously. While the documented behavior aligns with the stated purpose of task monitoring and automated alerting, the requirement for high-privilege credentials and external network communication for data reporting constitutes high-risk capabilities as defined in the review criteria (SKILL.md, README.md).
Capability Assessment
Purpose & Capability
The declared purpose (DataWorks monitoring, alerting, report delivery) is coherent with the actions described in SKILL.md. However, the skill expects DataWorks ACCESS_KEY_ID/ACCESS_KEY_SECRET and Feishu webhook configuration (per README and SKILL.md's TOOLS.md guidance) but the registry metadata lists no required environment variables or credentials — a clear mismatch.
Instruction Scope
SKILL.md instructs spawning asynchronous child agents, calling DataWorks APIs to fetch task instances and error logs, running LLM-based analysis, and sending detailed reports (including error logs) to Feishu and @-mentioning users. Those steps are within the stated monitoring purpose but include handling and transmitting potentially sensitive logs and using external endpoints; the instructions lack specifics about credential handling, redaction/desensitization steps, and access controls.
Install Mechanism
No install spec and no code files beyond documentation — lowest-risk install model. The skill is instruction-only, so nothing is automatically downloaded or written to disk by an installer in the manifest.
Credentials
Functionality clearly requires sensitive credentials (DataWorks access keys) and Feishu webhook/API credentials, but the skill does not declare these required env vars or a primary credential. Requesting unlisted secrets (or expecting them to be placed in TOOLS.md) is disproportionate and obscures the scope of access needed.
Persistence & Privilege
The skill is not always-enabled and does not request system-wide persistence. It spawns child agents (described in SKILL.md) to run asynchronously — that matches the stated asynchronous design and is not flagged on its own. No modifications to other skills or global configs are described.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install dataworks-smart-monitor - After installation, invoke the skill by name or use
/dataworks-smart-monitor - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release of dataworks-smart-monitor.
- Monitors DataWorks task statuses and analyzes failures.
- Provides intelligent alert classification (P0/P1/P2).
- Generates daily reports with task summaries.
- Notifications for severe and important alerts.
- Synchronous execution in main session for this version.
Metadata
Frequently Asked Questions
What is Dataworks Smart Monitor?
DataWorks 智能监控技能 - 异步分析任务运行情况,智能告警分级(不阻塞主会话). It is an AI Agent Skill for Claude Code / OpenClaw, with 109 downloads so far.
How do I install Dataworks Smart Monitor?
Run "/install dataworks-smart-monitor" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Dataworks Smart Monitor free?
Yes, Dataworks Smart Monitor is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Dataworks Smart Monitor support?
Dataworks Smart Monitor is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Dataworks Smart Monitor?
It is built and maintained by alexmayanjun-collab (@alexmayanjun-collab); the current version is v1.0.0.
More Skills