mcpField
F
feature-flag-rollout-review

feature-flag-rollout-review

github.com/sylphxaiCloud & DevOps145 installsUpdated today
Repository

About feature-flag-rollout-review

A skill for structuring risky feature-flag rollouts into plans with owners, gates, metrics, and rollback paths. It distinguishes flag types and covers targeting, monitoring, support, and cleanup.

Key features

  1. 01Identifies rollout risks and dependencies
  2. 02Distinguishes six feature-flag types
  3. 03Defines targeting, metrics, and rollout gates
  4. 04Creates state machines, decision tables, and cleanup lists

Use cases

  1. 01Planning a staged feature release
  2. 02Reviewing rollback and monitoring readiness
  3. 03Organizing ownership and cleanup for flags

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/sylphxai/skills --skill feature-flag-rollout-review

After installing, type/feature-flag-rollout-reviewin Claude Code to confirm it loaded.

Bundled filesagents/openai.yamlreferences/feature-flag-rollout-patterns.md

Details

From the project README

Feature Flag Rollout Review

Use this skill to convert a risky product, operations, trust, or marketplace question into a concrete artifact with owners, gates, metrics, and recovery paths.

Workflow

  1. Identify feature owner, risk class, user segments, dependencies, exposure unit, rollout channel, and rollback mechanism.
  2. Read references/feature-flag-rollout-patterns.md.
  3. Separate release flag, experiment flag, permission flag, ops kill switch, migration flag, and configuration flag.
  4. Define targeting, metrics, guardrails, rollout gates, support/comms, and cleanup deadline.
  5. Produce flag plan, rollout state machine, decision table, event schema, and cleanup checklist.

When not to use

  • Do not use for generic advice the base model already handles without this skill's specific artifact contract.

Guardrails

  • Do not use feature flags as permanent product architecture without ownership and cleanup.
  • Do not target sensitive groups in ways that create unfairness, privacy risk, or unsupported claims.
  • Do not roll forward after guardrails fail just because deploy checks are green.
  • Every high-risk flag needs an owner, rollback path, monitoring, and support note.

Output format

Feature flag context:
Risk / owner / exposure unit:

Rollout plan:
| Stage | Audience | Gate metric | Guardrail | Decision | Owner |
| --- | --- | --- | --- | --- | --- |

Flag inventory:
- <flag> -> <type, default, dependencies, cleanup date>

Rollback and cleanup:
- <trigger> -> <disable/rollback/comms/remove>

Comments

Markdown supported