Score growth experiments using the ICE framework and update experiment JSON.
npx skills add https://github.com/StreamPilotOrg/growth-machine --skill ice-scorer
Please help me install this Agent Skill. Skill: ice-scorer Source: https://github.com/StreamPilotOrg/growth-machine/tree/main/.claude/skills/ice-scorer Command: npx skills add https://github.com/StreamPilotOrg/growth-machine --skill ice-scorer
This Skill helps you quickly evaluate and rank growth experiments using the ICE framework, removing guesswork and aligning prioritization with business impact.
When you create a new experiment or ask to prioritize, invoke the ICE Scorer to generate a preliminary score or to re-score backlog experiments. Provide the hypotheses and any available data; the skill will return an ICE score and a recommended priority.
High-intent search queries and answers about installing and using this skill.
The ICE framework prioritizes experiments by scoring Impact, Confidence, and Ease on 1-10 scales, then multiplying them for a total score. Higher scores indicate experiments worth running first, removing guesswork from prioritization decisions.
ICE scoring measures Impact (business value), Confidence (likelihood of success), and Ease (implementation difficulty) on 1-10 scales. The total is their product; scores of 300 or higher are recommended for execution.
Use ICE scoring when creating new experiments, comparing multiple candidates, deciding which experiment to run first, or analyzing a backlog to identify high-value opportunities and align prioritization with business impact.
Yes, you can rescore experiments at any time to reflect new data, changing conditions, or refined estimates. The updated scores help you re-rank your backlog and adjust execution order.
When an experiment's ICE total meets or exceeds 300, it is routed to the pipeline for execution, ensuring high-impact opportunities move forward consistently.
Provide the experiment hypothesis and any available supporting data. The Skill then assesses Impact, Confidence, and Ease, delivering a justified score and priority recommendation.
FunnelEnvyGenerate testable CRO hypotheses and a prioritized experiment roadmap from L0/L1 context.
DavidSchmidt00Calculate RICE scores for product features using user-provided estimates.
finsilabsAudit e-commerce checkout drop-off and prioritize A/B tests with ICE scoring.
autostay-krCompare nine product prioritization frameworks including RICE, ICE, Opportunity Score, and MoSCoW with formulas and templates.
