prometheus-skill

Name: prometheus-skill
Rating: 7.5 (25 reviews)
Author: XiaoMi

by XiaoMi1.1k178GitHub

This skill provides a structured 7-step process for diagnosing service issues using Prometheus metrics. It queries CPU, memory, load, and JVM heap metrics, then analyzes the results to identify potential problems. The instructions include specific PromQL queries and example outputs with interpretations.

Output Preview

output_preview.md

Service Diagnostic Report: xxx_gis

📊 Resource Utilization Summary

CPU Metrics

CPU Usage: 32.35% (of requested cores)
CPU Requested: 6181% (61.81 cores)
Load Average (10s): 1242.0
Analysis: CPU utilization is healthy at 32.35% of requested capacity. Load (1242) is well below 80% of CPU request (4944.8), indicating no CPU saturation.

Memory Metrics

Memory Usage: 21.61 GB
Memory Limit: 75.16 GB
Utilization: 28.75%
Analysis: Memory usage is within safe limits at 28.75% of allocated memory. No immediate memory pressure detected.

JVM Heap Metrics

Heap Used: 15.09 GB
Heap Max: 100.50 GB
Heap Utilization: 15.01%
Analysis: Heap utilization is excellent at only 15.01%. No signs of memory leaks or garbage collection issues.

🎯 Overall Assessment

Status: ✅ HEALTHY Confidence: High

Key Findings:

All resource metrics are within optimal ranges
CPU load is significantly below capacity threshold
Memory and heap utilization show ample headroom
No immediate action required

Recommendations:

Continue monitoring as normal
Consider right-sizing if consistent low utilization persists
Next check scheduled: 24 hours

🔍 Raw Metric Values

{
  "cpu_usage_percent": 32.3487092130255,
  "cpu_requested_percent": 6181,
  "memory_usage_bytes": 21611335680,
  "memory_limit_bytes": 75161927680,
  "load_average_10s": 1242.0,
  "heap_used_bytes": 15085025864,
  "heap_max_bytes": 100503912448,
  "timestamp": "2024-01-15T10:30:45Z",
  "application": "xxx_gis"
}

Target Audience

DevOps engineers and SREs who need to quickly diagnose service issues using Prometheus metrics in their monitoring stack

10/10Security

Low security risk, safe to use

Clarity

Practicality

Quality

Maintainability

Innovation

DevOps

prometheusmonitoringservice-diagnosticsmetrics-query

Compatible Agents

Claude Code

~/.claude/skills/

Codex CLI

~/.codex/skills/

Gemini CLI

~/.gemini/skills/

OpenCode

~/.opencode/skills/

OpenClaw

~/.openclaw/skills/

GitHub Copilot

~/.copilot/skills/

Cursor

~/.cursor/skills/

Windsurf

~/.codeium/windsurf/skills/

Cline

~/.cline/skills/

Roo Code

~/.roo/skills/

Kiro

~/.kiro/skills/

Junie

~/.junie/skills/

Augment Code

~/.augment/skills/

Warp

~/.warp/skills/

Goose

~/.config/goose/skills/

SKILL.md

基础规则

当用户询问类似 "我的id:xxx name:gis 服务现在什么问题？"、"xxx_gis服务有啥问题没" 或 "帮我看下projectId=xxx, projectName=gis 的服务现状" 等想查询服务监控指标的问题时，请按照以下 7 个步骤进行排查，并将结果汇总分析服务的当前状态(注意替换application)。

查询当前 CPU 使用量 构造 promql 语句并运行 prometheus.py，例如：
```
uv run prometheus.py 'sum(rate(container_cpu_user_seconds_total{image!="",application="xxx_gis"}[30s])) by(application) * 100'
```
结果举例："查询到的最新值: 32.3487092130255" 代表查询使用了 32% 左右的 CPU（最大不是 100%，而是申请核心数的百分比，如申请 3 核心则最大是 300%）。
查询当前 CPU 申请量 构造 promql 语句并运行 prometheus.py，例如：
```
uv run prometheus.py 'sum((container_spec_cpu_quota{application="xxx_gis", image!=""}) /1000) by (application)'
```
结果举例："查询到的最新值: 6181" 代表这个服务总共申请了 6181% 的 CPU。
查询当前 Memory 使用量 构造 promql 语句并运行 prometheus.py，例如：
```
uv run prometheus.py 'sum(container_memory_rss{image!="",application="xxx_gis"}) by (application)'
```
结果举例："查询到的最新值: 21611335680.0" (单位是 byte)
查询当前 Memory 申请量 构造 promql 语句并运行 prometheus.py，例如：
```
uv run prometheus.py 'sum(container_spec_memory_limit_bytes{image!="",application="xxx_gis"}) by (application)'
```
结果举例："查询到的最新值: 75161927680.0" (单位是 byte)
查询当前 Load 构造 promql 语句并运行 prometheus.py，例如：
```
uv run prometheus.py 'sum(container_cpu_load_average_10s{application="xxx_gis"}) by (application)'
```
结果举例："查询到的最新值: 1242.0" ,注意这个值应该和cpu申请量对比看，理论上小于申请量的80%为最佳，比如cpu申请量是2000，则1242并不高
查询当前 HeapUsed 使用量 构造 promql 语句并运行 prometheus.py，例如：
```
uv run prometheus.py 'sum(jvm_memory_used_bytes{application="xxx_gis"}) by (application)'
```
结果举例："查询到的最新值: 15085025864.0" (单位 byte)
查询当前 HeapMax 最大量 构造 promql 语句并运行 prometheus.py，例如：
```
uv run prometheus.py 'sum(jvm_memory_max_bytes{application="xxx_gis"}) by (application)'
```
结果举例："查询到的最新值: 100503912448.0" (单位 byte)

Source: https://github.com/XiaoMi/mone#jcommon~skills~prometheus-skill

Content curated from original sources, copyright belongs to authors

Grade A

7.5AI Score

Best Practices

Checking...

Try this Skill

User Rating

USER RATING

0UP

0DOWN

Loading files...

WORKS WITH

Claude

Codex

Gemini