Feature Flag Lifecycle Auditor

The problem

Feature flags often evolve into permanent technical debt, causing server/client evaluation drift or dangerous "fail-open" behaviors during SDK outages. Managing stale experiments and ensuring data compatibility during rollbacks is difficult to track across complex codebases.

What it does

• Identifies stale flags, ownerless toggles, and expired experiments that are ready for cleanup.
• Audits evaluation consistency across server, browser, and mobile environments to prevent tenant leakage.
• Evaluates kill-switch effectiveness and fallback safety for high-risk features.
• Maps flag dependencies and nested evaluations to prevent state explosion and circular logic.
• Generates a prioritized cleanup plan with a 100-point lifecycle health score.

Why this beats prompting it yourself

General LLMs often miss the subtle data compatibility risks of a flag rollback or the difference between a release toggle and an authorization gate. This skill uses a specialized rubric to detect side effects like duplicate payments or API contract breaks that simple code reviews overlook.

Use cases

• Auditing a legacy flag inventory before a major infrastructure migration.
• Reviewing a complex percentage rollout plan for cross-platform consistency.
• Identifying dead code branches and stale experiments for a sprint cleanup task.
• Verifying that new feature toggles have safe fallbacks and proper observability hooks.

Known limitations

Does not perform live production changes or alter targeting rules in your provider dashboard. Requires manual input of flag configurations, SDK wrappers, or audit logs for analysis.