用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt;运行前请先检查来源。
npx skills add https://github.com/lev-os/agents --skill five-whys命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地?可下载 SkillsMP 当前能够提供的文件。
Routes adaptive multi-agent deliberation with fractal context cycles. Use when using /cdo, think/deep/debug/parliament work, or long runs paired with autoresearch scheduling.
Use when you need an evidence-first provenance report from `cass`, repo docs, `.lev`, `~/.agents/diagrams`, `qmd`, or Grep/Glob tools.
Use the codebase knowledge graph for structural code queries. Triggers on: explore the codebase, understand the architecture, what functions exist, show me the structure, who calls this function, what does X call, trace the call chain, find callers of, show dependencies, impact analysis, dead code, unused functions, high fan-out, refactor candidates, code quality audit, graph query syntax, Cypher query examples, edge types, how to use search_graph.
基于 SOC 职业分类
正在显示 SKILL.md
| name | five-whys |
| description | Uncover root causes by iteratively asking why a problem occurred, moving from symptoms to underlying system failures |
Developed by Sakichi Toyoda and Taiichi Ohno at Toyota, the Five Whys is a root cause analysis technique that asks "why?" five times (or until reaching root cause) to move from problem symptoms to underlying system failures. Each answer forms the basis of the next "why?" The method reveals that surface problems (machine stopped) often trace to systemic issues (no maintenance strainer).
Write a specific, observable problem statement. Avoid vague language. Bad: "Quality is poor." Good: "Product X failed QA inspection 3 times this week due to misaligned components."
Example: "The website went down for 2 hours on Tuesday afternoon, affecting 10,000 users."
Gather those directly involved in the process—they have the detailed knowledge. Don't rely solely on management perspectives removed from day-to-day reality.
Example: Include the on-call engineer who responded, the developer who deployed, and the QA tester who reviewed the change.
Ask why the problem occurred. Each answer becomes the input for the next "why?" Don't accept vague answers—push for specific, verifiable causes.
Example:
Continue asking until reaching a root cause you can act on. Signs you've hit root: answer is a process gap, missing system, or design flaw—not a person or one-time event.
Example: Root cause is "No onboarding process for distributed systems concepts," not "Developer made a mistake" (blame) or "Connection leaked" (symptom).
Fix the root cause, not the symptom. Verify the fix prevents recurrence, not just the specific instance.
Example: Countermeasure: Create distributed systems onboarding with connection pooling best practices, add automated tests for resource leaks, update code review checklist. Result: No further connection pool failures.
Problem: Machine stopped working
Root Cause: Missing strainer (design/process flaw) Countermeasure: Install strainer on all lubrication pumps
Situation: E-commerce company experiences spike in customer service complaints about late deliveries.
Application:
Root Cause: Missing infrastructure (charging station), not "lazy workers"
Countermeasure: Install 5 charging stations in returns area, add scanner battery checks to opening checklist, train returns team on scan importance
Outcome: Late deliveries drop 60%, complaints fall 75%, warehouse productivity increases 20%.