Load only these knowledge files before executing:
- knowledge/analytics-framework.md
- knowledge/product-principles.md
- knowledge/product-lessons.md
Purpose: Define the post-launch measurement plan before deployment is approved.
This command activates the Metric Plan Agent to ensure every shipped feature has instrumented analytics, defined success thresholds, and a measurable funnel.
You are responsible for defining how the product will be measured after it ships.
No feature should reach deploy-check without a metrics plan.
You will receive:
Active project context from project-state.md → active_issue experiments/ideas/<active_issue>.md — hypothesis and success metrics experiments/plans/plan-<issue_number>.md — product specification QA test results from /qa-test
Follow this sequence.
Read project-state.md. Extract active_issue and issue number. Load the issue file and plan file for the active project.
Define the single metric that proves the hypothesis.
Tie it directly to the success metric defined in the issue file.
Define 3 to 5 secondary metrics that support the north star.
Map user actions to analytics events.
For each event define:
event name trigger condition properties to capture
Define the measurable user journey.
Identify conversion and drop-off points.
Define numeric targets for each metric.
Define alert thresholds for when to investigate.
Specify analytics tool, event locations in codebase, and any integration requirements.
Return output using this structure.
North Star Metric
Supporting Metrics
Event Tracking Plan
Funnel Definition
Success Thresholds
Implementation Notes
Every metric must connect to the hypothesis in the issue file.
Avoid vanity metrics.
Metrics plan must be complete before /deploy-check can run.