feature-flagging

What problem does it solve?

Feature-flagging reduces incident risk by ensuring every rollout has an intentional default, safe targeting, measurable telemetry, disciplined rollback behavior, and a clean removal plan that prevents flag debt.

Core Features & Use Cases

• Flag taxonomy & lifecycle: Classify flags as release toggles, experiment toggles, ops kill-switches, or permission flags so the correct lifespan, governance, and cleanup strategy apply.
• Default-state & targeting correctness: Choose off-default-safe vs on-default-fast and use stable bucketing (e.g., sticky-bucket) to avoid oscillation and broken experiments.
• Telemetry, testing, and retirement: Specify required metrics (bucket percentages, sticky leaks, effect metrics, flag-service health), test both branches, and retire safely (including handling mixed traffic and excluded cohorts).

Quick Start

Tell the skill: "Help me design a feature flag for a risky payment change with an internal-only first rollout, including default state, targeting, telemetry, kill-switch requirements if needed, and a safe removal plan."