可觀測性
可觀測性
規劃可回答維運問題的 logs、metrics 與 tracing。
Community · 社群來源 · DevOps 與基礎架構
來源狀態
可用來源
原始名稱: observability-and-instrumentation
原始作者: Addy Osmani
這是第三方 Community Skill;2lus 收錄不代表已完整安全審計或保證安全。
資源類型
Skill
與類似 Skill 有什麼不同?
- 系統化除錯
Observability 建立 logs、metrics 與 traces 以提供證據;Systematic Debugging 使用可得證據調查特定故障。
來源描述的能力
這些是來源描述的可能操作,不代表 2lus 已授予權限或已測試。
尚未列出能力,請審查原始來源。
這個 Skill 是什麼?
讓服務行為留下可關聯且成本可控的證據。
可以做什麼?何時適合使用?
- 補足請求追蹤
- 定義錯誤率與延遲指標
- 減少無法處理的告警
如何使用
- 列出值班人員需要回答的問題
- 將問題對應事件、指標與 span
- 設計 correlation ID、遮蔽與告警條件
使用前你需要準備
- Dockerfile 或 workflow
- 服務依賴與部署限制
你可以替換:
[service]、[environment]
環境與相依需求
- 需要服務邊界、現有 telemetry 與維運目標;保留來源參考。
使用範例與 Prompt
以下是 2lus 撰寫的示範需求;請替換為你有權處理的檔案與專案,不代表已執行或保證結果。
入門
為背景匯出任務設計開始、完成、失敗事件及 correlation ID,不連線監控平台。
實務
提出 API 的 rate、errors、duration 與跨服務 span 設計,註明 health check 與告警用途。
進階
替重試佇列設計 SLI/SLO 對應、sampling、指標 cardinality 上限與可操作告警;列出保留期限及個資遮蔽規則。
實用提醒
- 避免高 cardinality 指標標籤,記錄可查詢的穩定事件名稱。
限制與注意事項
- 不取代即時 incident response,也不代表已建置監控平台。
安全注意事項
- 禁止完整記錄 token、密碼或請求內容;採欄位白名單。
使用與設定
此條目不提供已確認的通用安裝指令;請依官方文件與 Agent 版本操作。
支援平台
未確認特定 Agent 相容性
來源與授權
來源查核日期(非安全認證): 2026-09-29
MIT
原始來源 ↗ 授權條款 ↗ 官方文件 ↗相關 Skills
使用第三方 Skill 前,請先檢查來源、權限與執行內容。安裝指令只供查看與複製,不會由 2lus 執行。
2lus AI Skills Library 提供 Skill 的整理與使用導覽。第三方 Skill 的內容、授權與可用性以原始來源為準。使用或安裝前,請自行確認其權限與執行內容。
Observability
Plan logs, metrics and traces that answer operational questions.
Community · Community source · DevOps & Infrastructure
Source status
Active source
Original name: observability-and-instrumentation
Original author: Addy Osmani
This is a third-party community skill. Inclusion by 2lus is not a complete security audit or safety guarantee.
Resource type
Skill
How is this different from similar skills?
- Systematic Debugging
Observability instruments logs, metrics and traces; Systematic Debugging uses available evidence to investigate a specific failure.
Documented capabilities
These are operations described upstream, not permissions granted or tested by 2lus.
Capabilities not declared here; review the original source.
What is this skill?
Produce correlated evidence with controlled telemetry cost.
Use cases and when to use it
- Fill request tracing gaps
- Define error and latency signals
- Reduce unactionable alerts
How to use it
- List on-call questions
- Map questions to events, metrics and spans
- Specify correlation, redaction and alerts
What you need
- Dockerfile or workflow
- Service dependencies and deployment constraints
You can replace:
[service], [environment]
Environment and dependencies
- Provide service boundaries, existing telemetry and operational goals; retain references.
Usage and prompt examples
These example requests were written by 2lus. Substitute files and projects you may use; examples are not executed results or guarantees.
Beginner
Design started, completed and failed export-job events with correlation IDs without connecting to monitoring.
Practical
Propose API rate, error and duration signals plus cross-service spans; explain health checks and alert use.
Advanced
Design retry-queue SLI/SLO signals, sampling, cardinality limits and actionable alerts with retention and redaction policies.
Tips
- Avoid high-cardinality labels and prefer stable event names.
Limitations
- This is instrumentation design, not live incident response or a hosted monitoring service.
Security notes
- Allowlist fields; never record tokens, passwords or full request bodies.
Usage and setup
No verified universal installation command is provided for this entry. Follow the official documentation for your agent version.
Supported agents
Specific agent compatibility unknown
Sources and license
Source check date (not a safety certification): 2026-09-29
MIT
Original source ↗ License terms ↗ Documentation ↗Related skills
Before using a third-party skill, review its source, permissions and executable content. Commands are for viewing and copying only; 2lus does not execute them.
2lus AI Skills Library provides curated educational guides. Third-party content, licenses and availability are governed by their original sources. Review permissions and executable content before use or installation.