More screenshots
Works with the AI Agents you already use
Feature Flag Retirement Auditor
Delete stale flags without stranding cohorts or losing rollback.
$19.99
Feature Flag Retirement Auditor
Example session with this skill installed
Audit the retirement of checkout_v2. Production defaults to true and the web dashboard reports 100% rollout. The server registry contains the flag, an admin service is the only known writer, and both checkout branches still exist. Older mobile versions evaluate the flag locally. Evaluation caches have a five-minute TTL. A revenue dashboard groups by the flag variant. Rollback after deleting the configuration has not been defined.
- Read your context and instructions
- Compiled the feature flag retirement
Verdict — STAGED-RETIREMENT
The terminal branch is the enabled checkout path, but retirement is not ready. The visible web cohort has converged while old mobile evaluation, analytics continuity, and post-removal rollback remain unresolved.
Flag identity
Canonical key: checkout_v2. Known definition: server registry. Known writer: admin service. Production default: true. Terminal behavior: enabled branch. Aliases and environment-specific variants were not supplied and remain an explicit search requirement.
Ten-domain coverage
| Domain | Status | Evidence |
|---|---|---|
| Definition | covered | server registry entry |
| Writers | covered | admin service is the only supplied writer |
| Readers | gap | older mobile local evaluator unresolved |
| Defaults | covered | production default true |
| Cohorts | gap | mobile-version distribution not segmented |
| Caches | covered | five-minute evaluation TTL |
| Analytics | covered, cleanup needed | dashboard groups by variant |
| Tests | covered | both branches remain represented |
| Documentation | covered | runbook reference identified |
| Rollback | gap | recovery after configuration deletion undefined |
Staged retirement plan
- Resolve aliases and the oldest supported mobile evaluator.
- Segment convergence by app version, tenant, region, and offline-client behavior.
- Freeze the terminal value and stop all writes.
- Wait through cache expiry and verify no cohort changes branch.
- Remove runtime evaluations while keeping an emergency code rollback.
- Delete the disabled branch and update tests.
- Migrate dashboards away from the variant dimension without breaking historical comparisons.
- Update documentation and remove remote configuration last.
Discriminating tests
Replay the oldest supported mobile client, test evaluation during and after cache expiry, verify dashboard continuity, and drill rollback using a release artifact that does not depend on the deleted flag record.
Validator
All ten coverage domains are present and the record passes the bundled structural validator. No flag, cache, application, dashboard, or production environment was modified.
Connects securely to your tools. The creator never sees your data.
About this skill
A flag is distributed state, not one conditional
Definitions, writers, readers, defaults, remote evaluation, old clients, cohorts, caches, analytics, tests, documentation, and rollback can keep a feature flag alive after a rollout appears complete.
Feature Flag Retirement Auditor resolves one canonical identity, inventories ten lifecycle domains, proves cohort convergence, selects the terminal branch, and sequences a reversible cleanup. Configuration is removed only after evaluations, dead behavior, analytics, tests, documentation, and recovery have been reconciled.
Built for platform, product-infrastructure, mobile, and release engineers cleaning up rollout flags, experiment switches, operational toggles, and kill switches. It profiles the real flag SDK, evaluation location, client lifecycle, analytics naming, cache policy, tests, and deployment model.
The package includes full Skill source, a ten-domain coverage method, worked checkout example, deterministic validator, mutation tests, and a six-page Buyer Guide PDF.
Boundaries
It does not flip flags, edit code, deploy, or infer that a platform-wide percentage represents every tenant, region, version, or offline client.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse All Claude Skills.
Requires buyer-supplied flag definitions, search results, cohort reports, tests, dashboards, or rollout notes. It does not flip flags, edit code, deploy, or infer that a platform-wide percentage represents every tenant, region, version, or offline client.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 2 months ago
Permissions
No special permissions declared or detected