SkillVaultskills Browse all 500 skills

Maintenance · Version 1.3.0 · Reviewed 2026-08-02

Feature Toggle Cleanup Advisor

Reduce change risk for stale flag detection and branch removal safety with evidence, explicit trade-offs, and a verification plan.

4 method steps 4 documented failure modes 4 diagnostic checks 7 quality gates

Removes stale toggles safely by proving the dead branch is unreachable before deleting it.

₹149 one-time

Get this skill archive

What this skill helps you do

  • Stale flag detection
  • Branch removal safety
  • Cleanup sequencing

How Feature Toggle Cleanup Advisor works

You provide

Current versions, changelogs, support dates, and coverage

It inspects

Breaking changes and transitive pins behind stale flag detection

It decides

A branch removal safety sequence taking one major at a time

You verify

Suite green between increments with rollback proven

What it checks first

Feature Toggle Cleanup Advisor removes stale toggles safely by proving the dead branch is unreachable before deleting it. Use it when the work involves Stale flag detection, Branch removal safety, Cleanup sequencing.

  1. Whether the current version is still supported, and the date support actually ends.
  2. How many major versions separate current from target, since each carries its own breaking changes.
  3. Whether test coverage is sufficient to detect a behavioral regression from the upgrade.
  4. Transitive dependencies that constrain the upgrade regardless of direct requirements.

Failure modes it recognizes

  • Skipping intermediate majors so several sets of breaking changes land together, unattributable.
  • A deprecation warning ignored until the removal makes the upgrade a project.
  • A transitive dependency pinning an old version, silently blocking the direct upgrade.
  • An upgrade completed without a rollback path because a data format changed.

Answers it will reject

  • Upgrading everything at once to reduce disruption, which maximizes debugging cost.
  • Treating a green build as sufficient evidence when coverage of the changed behavior is thin.
  • Deferring an end-of-life migration until support has already lapsed.

Decision rules it applies

  • Upgrade one major at a time with the suite green between each step.
  • Resolve deprecation warnings from version N before starting N+1; they are the removal list.
  • Plan end-of-life migrations against the support date, not against convenience.

Evidence it asks for

  • Inventory direct and transitive versions against their support timelines.
  • Run the suite between each increment and record which behavior changed.
  • Verify rollback by deploying the previous version against the upgraded data.

The method inside

  1. Map the artifact, actors, boundaries, and invariants relevant to stale flag detection.
  2. Trace concrete failure or abuse paths for branch removal safety; do not report checklist items without a mechanism.
  3. Prioritize cleanup sequencing findings by impact, likelihood, confidence, and cost of correction.
  4. Recommend the smallest defensible change, then define how an independent reviewer can verify it.

Deliverables

  • Stale flag detection assessment
  • Branch removal safety decision and action plan
  • Cleanup sequencing verification checklist

Evidence requirements

  • Current and target versions
  • Dependency graph and changelogs
  • Tests, compatibility constraints, and rollout environment

Quality gates

  • Every material claim traces to supplied evidence or is labeled as a hypothesis.
  • The response follows the declared deliverable contract.
  • No execution, access, measurement, or verification is invented.
  • Secrets and personal data are redacted rather than repeated.
  • The user receives a concrete independent verification step.
  • The relevant failure modes in this domain were considered rather than only the reported symptom.
  • No listed anti-pattern was recommended as a solution.

Example task

Input

We have 60 flags that appear permanently on. How do we remove the dead branches without breaking anything?

Expected output

Appearing on is not the same as being on everywhere, and that difference is where the outage comes from. Instrument evaluation by variant for a full business cycle, and remove only flags where the off branch recorded zero evaluations across all environments...

Boundaries and compatibility

Ideal for

  • Stale flag detection: produce a decision or artifact grounded in supplied evidence.
  • Branch removal safety: produce a decision or artifact grounded in supplied evidence.
  • Cleanup sequencing: produce a decision or artifact grounded in supplied evidence.

Out of scope

  • Blindly upgrading across multiple major versions
  • Assuming semantic versioning guarantees compatibility

Agent compatibility

  • GitHub Copilot custom agents
  • Claude Agent Skills / SKILL.md
  • Any instruction-following chat model

Tool policy: Advisory by default. No tools are assumed. If the host provides tools, use read-only evidence gathering unless the user explicitly approves a scoped write or execution action.