OKR & Metric Validity Audit: AI Skill for Product Management
OKR & Metric Validity Audit is an AI skill that tests every metric for vanity, falsifiability, and gaming exposure. Each finding names the violated principle, carries a severity rating, and comes with a concrete rewrite.
npx skills add Uxcel-Lab/product-skills --skill pm-okr-metric-validity-audit
What this skill does
Tests OKRs, KPIs, and North Star Metrics for validity instead of polishing their wording.
- Rewrites output-counting key results as the behavior change they should cause.
- Tests every key result for falsifiability: baseline, target, unit, and timeframe.
- Exposes vanity metrics by asking what decision changes when the number moves.
- Checks objectives inspire direction instead of listing tasks in corporate fog.
- Traces each metric up the hierarchy to user value and business success.
- Asks how a cynical team could game each target, then names the missing guardrail.
- Balances lagging outcomes with leading indicators that allow course-correction.
- Rates findings by decision damage and ends with the top three rewrites.
When to use it
Use it when you need an AI assistant to:
- audit quarterly OKRs before the team commits to them;
- check a North Star Metric connects user value to business success;
- sweep a dashboard for vanity and orphan metrics;
- add guardrails before a growth target goes live;
- rewrite unfalsifiable key results into measurable ones;
- validate metrics defined in a spec, roadmap, or strategy doc.
What this audit checks
| Category | What it catches |
|---|---|
| Outcome vs. output | Key results that count launches, tickets, and shipped features |
| Falsifiability | Key results with no baseline, target, unit, or timeframe |
| Vanity exposure | Downloads, page views, and cumulative totals that inform no decision |
| Objective quality | Task-list objectives, corporate fog, too many objectives per cycle |
| Causal connection | Orphan metrics with no link to user value or the North Star |
| Gaming & guardrails | Targets a cynical team could hit at the product's expense |
| Leading/lagging balance | All-lagging metric sets that report failure only after the fact |
How the skill works
1
Establish context
The skill confirms the business model, product stage, and which metric level it is auditing.
2
Test every metric
It checks each objective and key result for outputs, vanity, falsifiability, gaming, and value connection.
3
Deliver ranked rewrites
It rates findings by decision damage and ends with the top three fixes, each with a rewrite.
Try it with
Related skills
1.0.0
MIT
Product Management
Audits
