ever-works/orgs/root-cause-drill
Claude Code and agent skill by ever-works/orgs
- community
Checking that the SKILL.md file is still in the repository.
Use when a problem keeps recurring or a symptom is being treated instead of a cause — drills through the causal chain with disciplined "why" iterations until an actionable root is reached.
Install
npx vskill@latest install ever-works/orgs/root-cause-drillSKILL.md
Root-Cause Drill
Follow the causal chain past the symptom. The failure mode this prevents: fixing the visible breakage repeatedly because the generating condition — a process gap, a missing owner, a bad default — was never touched.
Steps
- Write the problem as an observed event with time and impact ("the report shipped with stale numbers on the 12th, client caught it"), never as a judgment ("QA is sloppy").
- Ask why it happened and write the answer as a fact you can verify, not a guess. Verify it before drilling further — a chain built on an unverified link is fiction.
- Repeat the why on each answer. Five iterations is the folk rule; the real stop condition is reaching a cause that satisfies all three:
- Actionable: someone could change it with a decision or a process edit.
- Generative: fixing it prevents the class of problem, not just this instance.
- Within reach: it belongs to the organization, not to human nature or the weather.
- Branch when a why has two genuine parents (it usually does once): follow each branch separately rather than averaging them into mush. Most incidents have 2-3 roots.
- Watch for the two classic derailments:
- Blame arrival: the chain stops at a person. Push one more why — "why did the process let that mistake through?" People are how causes express themselves, not causes.
- Abstraction escape: answers drift to "lack of communication" or "culture". Re-anchor to the specific event and its verifiable facts.
- For each root, write the countermeasure at the root's level: a changed default, an added checkpoint, a reassigned ownership — with an owner and a date.
- Record how this class of problem will be detected sooner if the countermeasure leaks.
Quality bar
- Every link in the chain is a verified fact; mark any unverified link explicitly and treat conclusions past it as provisional.
- No chain ends at a person or at an abstraction.
- Each countermeasure maps to a specific root — a countermeasure with no root is a pet project smuggled in.
Output format
The event statement → the why-chain (with branches) showing verification status per link → roots that pass the three-part stop test → countermeasures with owners and dates → early-detection note.
Security
- Security scanPassed, 90 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingcompanies/metaframe-agency/skills/root-cause-drill/SKILL.md
Popularity
- Repository stars
- 2
- 7-day trend score
- 10
Stars and forks are for the whole GitHub repository, not this skill alone. Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/ever-works/orgs/root-cause-drill)