feature-flag-rollout-planner

When to invoke

• You are planning a safe rollout for a risky change.
• You want a consistent rollout template for engineering and product teams.

Inputs needed

• Feature context: name, description, target users.
• Risk profile: data migrations, performance risk, external dependencies.
• Success metrics and guardrail metrics.
• Optional stakeholders and comms channels.

Workflow

1. Validate inputs and normalize percentages/dates.
2. Propose rollout phases (internal, 1%, 10%, 50%, 100%) and duration.
3. For each phase, define:
   • enablement criteria
   • monitoring checklist
   • guardrails and rollback triggers
4. Produce comms checklist and ownership.

Output format

• Overview
• Phased rollout table
• Monitoring and dashboards
• Rollback plan
• Communications checklist

Guardrails

• If metrics are not provided, output placeholders instead of guessing.
• Include a default rollback trigger for error-rate regression.

Reference code

• plan_rollout.py

  reads a JSON feature brief and writes Markdown rollout plan.