Skip to main content
skill CX Codex

causal-inference-root-cause — Claude Code skill: root cause analysis

causal-inference-root-cause — a skill that pushes an agent through a five-step causal investigation, so the answer to why something happened comes with a mechanism and a tested alternative.

no license 110 installs ★ 203 updated 2026-09-01

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add lyndonkl/claude --skill causal-inference-root-cause -a claude-code
Open on GitHub

What causal-inference-root-cause does

The skill runs an agent through five steps: pin down the effect and its baseline, list competing explanations, build the causal model, test it, then write up and validate. Simple investigations get resources/template.md; harder ones are pointed at resources/methodology.md. Hypotheses come from repeated why-questions, cause categories, a timeline of what changed and differential diagnosis.

Testing borrows from epidemiology. Evidence is ranked from randomised trials down to expert opinion, and each link is checked against the 9 Bradford Hill factors, among them temporality, dose-response and plausibility. Confounders are mapped out in the open, and a correlation is not enough — a mechanism has to be named.

The write-up lands in causal-inference-root-cause.md and gets scored against resources/evaluators/rubric_causal_inference_root_cause.json. A 3.5 minimum is where the skill stops: below it, a tidy narrative without a counterfactual check still counts as unfinished.

What it helps with

  • trace a conversion drop back to a payment provider change
  • rule out the last deploy as the cause of an outage
  • test whether a metric shift is seasonal
  • write a postmortem that names a mechanism, not just symptoms
  • estimate the effect of a new policy against confounders

Example prompts for the agent

  • Find the root cause of the failing builds and show me the causal chain, not a list of correlations.
  • Run a causal analysis of the signup drop and test the competing explanations against each other.
  • Why did latency spike? Check the confounders and tell me what the counterfactual looks like.

How to install causal-inference-root-cause

1Skills CLI (Vercel)

npx skills add lyndonkl/claude --skill causal-inference-root-cause -a claude-code

2Manually: put the files in place

~/.claude/skills/causal-inference-root-cause/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/causal-inference-root-cause/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about causal-inference-root-cause

How do I install causal-inference-root-cause in Claude Code?

Run npx skills add lyndonkl/claude --skill causal-inference-root-cause -a claude-code. The skill file lands at ~/.claude/skills/causal-inference-root-cause/SKILL.md.

What does causal-inference-root-cause do?

It walks an agent through a five-step investigation of why something happened, keeping real causes apart from symptoms and spurious associations. You get a written analysis with a causal model, tested hypotheses and suggested interventions.

Which platforms can run this skill?

The same skill installs into Claude, Codex, Cursor, GitHub Copilot, Gemini CLI and OpenCode, each with its own install command and skill path. For Copilot the command is gh skill install lyndonkl/claude causal-inference-root-cause.

Similar skills

All skills for Science and research →

From the same set49

The whole set →
translation-reframing-audience-shift translation-reframing-audience-shift rewrites content for a different audience: technical docs to business cases, strategy to action steps, expert text to plain language. CX Codex +1 115installs meta-prompt-engineering The meta-prompt-engineering skill turns unclear prompts into structured instructions with a defined role, task steps, constraints, and quality checks. Use it when an agent’s responses vary between runs. CX Codex +1 114installs systems-thinking-leverage systems-thinking-leverage turns a tangled problem into a mapped system with ranked intervention points, built on feedback loops and Meadows' leverage ladder. CX Codex +1 114installs reviews-retros-reflection reviews-retros-reflection structures retrospectives, incident post-mortems, and periodic check-ins. Helps teams identify root causes and turn feedback into SMART action items without blame. CX Codex +1 113installs security-threat-model security-threat-model guides threat modeling across system components using STRIDE methodology, trust boundary maps, and defense-in-depth principles before audits or releases. CX Codex +1 109installs discovery-interviews-surveys discovery-interviews-surveys is a Claude Code skill for planning user interviews and surveys to test product assumptions and uncover customer problems, before building or when exploring why people switch products. CX Codex +1 109installs project-risk-register project-risk-register — a skill for identifying, assessing, and prioritizing project risks using probability-impact scoring. Assigns owners and mitigation plans, then tracks risk evolution throughout the project lifecycle. CX Codex +1 108installs forecast-premortem forecast-premortem is a Claude Code skill that stress-tests probabilistic forecasts by assuming failure before launch. It reveals overlooked risks and recalibrates overconfident estimates. CX Codex +1 108installs

Project sponsors

Companies that keep this analytics open