Staged Feature Rollouts

A flag inventory you can trust, a stage ladder with guardrail metrics and tested kill switches, weekly rollout reviews, drills and flag clean-up, for product teams and engineers releasing features gradually.

Free · Mac (Apple silicon) and Windows · v1.0.5 · Other platforms

Then open the Marketplace in the app, search for Staged Feature Rollouts, and click Install. It goes into your own files, and works offline from then on.

What it installs.

1

Pillar

1

Area

50

Projects

204

Starter tasks

🧩 Product Management

Staged Feature Rollouts

Releasing features gradually with flags, percentage rollouts and kill switches, watching metrics at each stage before going to everyone.

  1. Feature flag inventory with owners and expiry dates
  2. Choosing a feature flag platform
  3. Flag naming and lifecycle conventions
  4. Rollout stage ladder for the team
  5. Guardrail metrics baseline for rollouts
  6. Kill switches for critical user paths
  7. Rollout roles and decision rights
  8. Flag service outage and fallback defaults
  9. Targeting rules and consistent bucketing
  10. Weekly rollout review
  11. Stale flag clean-up rhythm
  12. Rollout calendar and freeze windows
  13. Stage gate checklist for every rollout
  14. Flag change audit log review
  15. Rollout dashboard split by flag variation
  16. On-call handover for active rollouts
  17. Quarterly kill switch drills
  18. Flag debt scorecard by team
  19. Rollout incident log and pattern review
  20. Detecting regressions at low traffic percentages
  21. Progressive delivery patterns study
  22. Separating deploy from release for product managers
  23. Testing both sides of every flag
  24. Rollout plan section for every spec
  25. Dwell time rules for each stage
  26. Database changes behind flags
  27. Automated halt on guardrail breach
  28. Server-side or client-side flag evaluation
  29. Migrating from config toggles to a flag service
  30. Account-level versus user-level rollouts
  31. Rollout status visible to support and sales
  32. Shortening time from merge to full release
  33. Flag permissions and production change approvals
  34. First percentage rollout of a new feature
  35. Staged rollout of a checkout or payment change
  36. Backend migration by traffic share
  37. Opt-in to default rollout of a redesign
  38. Rollouts around a peak trading period
  39. Blameless review of a halted rollout
  40. Staged rollouts for a small team without a platform
  41. Inheriting a codebase full of old flags
  42. Flags across web and app versions
  43. Rollouts for enterprise customers with change windows
  44. Teaching rollout practice to new engineers
  45. Rollout risk register for high-stakes releases
  46. Flag dependency and prerequisite map
  47. Regional and ring-based rollouts
  48. Shadow mode rollouts for model changes
  49. Release, experiment and permanent flag types
  50. Annual progressive delivery maturity review

Where it starts.

Each project arrives with its purpose, milestones and first tasks written out. Here is how the first three read.

Feature flag inventory with owners and expiry dates

This is a team plan, best installed into a shared workspace so everyone with a role in it can see and own their tasks. Most codebases that have used flags for a year carry dozens nobody can explain: a toggle from a launch two years ago, a percentage stuck at 50, a kill switch whose owner has left.

First tasks

  • Export the full flag list from the flag service or config files
  • Search the codebase for flag keys that do not appear in the export
  • Add purpose, owner, state and created date for every flag
  • Highlight flags held at a partial percentage for over 30 days
  • Refresh the flag inventory and reassign orphaned flags @recurring(yearly)

Choosing a feature flag platform

Homegrown toggles in a config file work until the team needs percentage rollouts, targeting by account, an audit trail and a kill switch that does not need a deploy. Comparing two or three platforms, including open source and self-hosted options, against the same requirements avoids buying on demo polish and finding out later that the SDK does not support your mobile client or your data location needs.

First tasks

  • List every language and client the flag SDK must support
  • Write the must-have and nice-to-have requirements on one page
  • Shortlist three platforms including one open source option
  • Run a one-week trial with a real flag in staging
  • Review flag platform cost and seat count before renewal @recurring(yearly)

Flag naming and lifecycle conventions

A flag called new_flow_v2 tells nobody what it controls, who owns it or when it should go. A one-page convention for names, descriptions, tags and lifecycle states, applied to every new flag from now on, makes search work, makes clean-up possible and makes the dashboard readable at two in the morning.

First tasks

  • Collect ten current flag names that confuse people and say why
  • Draft a naming pattern and the required fields for new flags
  • Agree the lifecycle states with the engineering leads
  • Add a link to the convention in the pull request template

Make it yours.

Install Staged Feature Rollouts, then tell Aurum the date you're aiming for. It checks whether the plan fits the weeks you actually have.

Free · Mac (Apple silicon) and Windows · v1.0.5 · Other platforms

Blueprints are general information to help you organise your own work. They are not medical, financial, legal or other professional advice. See the Terms.