design-kpis — Codex skill: metrics and installation
This skill helps define KPIs, drivers, guardrails, and measurement plans for product or business decisions. Use it when a team needs to choose what success means and how to track it.
npx skills add openai/plugins --skill design-kpis -a codexWhat design-kpis does
The skill starts by clarifying the decision and its context, then builds a candidate set of outcome, driver, and guardrail metrics. It helps compare candidates by relevance, usefulness for decisions, measurement cadence, team influence, and risk of misleading improvement. Recommendations can include calculation methods, data sources, tradeoffs, and caveats. Target setting is a separate step based on available evidence and assumptions.
When sources overlap or conflict, the workflow compares ownership, freshness, definitions, grain, and coverage, then checks selected data through live reads. If a required source cannot be accessed, the skill asks for it or a reviewed fallback. It uses $gather-business-context to collect business context; for reconciling existing metrics, it directs users to start with $analyze-data-quality.
Install it when you need to design or revise a KPI framework. For a task limited to reconciling existing metrics, start with $analyze-data-quality instead.
What it helps with
- choose KPIs for a product decision
- define how a metric is calculated and sourced
- select drivers that help diagnose KPI movement
- add guardrails when tradeoffs are likely
- set a target using evidence and stated assumptions
Example prompts for the agent
- Help me choose KPIs for a decision about launching this feature.
- Compare candidate retention metrics and suggest useful drivers and guardrails.
- Draft a measurement plan and explain what evidence we need to set a target.
How to install design-kpis
1Skills CLI (Vercel)
npx skills add openai/plugins --skill design-kpis -a codex2Manually: put the files in place
.agents/skills/design-kpis/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add openai/plugins --skill design-kpis -a claude-code2Manually: put the files in place
~/.claude/skills/design-kpis/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/design-kpis/
Manually: put the files in place
.cursor/skills/design-kpis/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install openai/plugins design-kpis2Manually: put the files in place
.github/skills/design-kpis/SKILL.mdPlatform CLI
gemini skills install https://github.com/openai/plugins --path plugins/data-analytics/skills/design-kpisManually: put the files in place
.devin/skills/design-kpis/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
From the same set201
The whole set →Similar
Questions about design-kpis
How do I install design-kpis in Codex?
Run npx skills add openai/plugins --skill design-kpis. You can also use npx skills add openai/plugins --skill design-kpis -a codex.
What does design-kpis do?
It helps define KPIs, drivers, and guardrails for product or business decisions, with metric definitions and a measurement plan. It treats target setting as a separate judgment based on evidence.
What should I do if metric sources conflict?
The workflow compares source ownership, freshness, definitions, grain, and coverage, then verifies selected data through live reads. If a required source is unavailable, it asks for that source or a reviewed fallback.
