Toggled

The Toggled blog

The Toggled blog

Practical release engineering: feature flags, gradual rollouts, kill switches and experiments. Written by the small team building Toggled.

· 9 min read

GitHub feature flags: what GitHub actually gives you, and what you have to build

GitHub runs one of the most sophisticated feature flag systems in the industry, and sells you none of it. Here is what you can genuinely gate with Actions alone, and the exact point where build-time configuration stops being enough.

Read the post

· 12 min read

Feature flag pricing models: what each meter actually costs you

Every vendor in this category charges for a different thing. The feature comparison is the easy part; the meter is what decides whether the tool is affordable for your specific shape of product.

Read the post

· 11 min read

How to roll out AI features safely with feature flags

An LLM feature fails differently from a normal one. It degrades rather than errors, the cost is unbounded, and the vendor can change the model underneath you. Flags are how you keep control of all three.

Read the post

· 11 min read

OpenFeature: what it is, and whether your team should actually adopt it

The pitch is that you can switch flag vendors by changing one line. That part is true. The question worth asking is whether the line of code was ever the expensive part of a migration.

Read the post

· 9 min read

Feature flags vs environment variables: when config stops being enough

An environment variable is a deploy-time decision. A feature flag is a runtime one. Most teams do not notice the difference until the night they need to change something without deploying.

Read the post

· 9 min read

Feature flag best practices: naming, ownership, and killing flag debt

Two of the four flag types are supposed to live forever, which is why "delete all the old flags" is advice that breaks production. Here is the system that works.

Read the post

· 8 min read

Unleash open source is end of life: what self-hosted teams should do next

Nothing switches off on January 1. The risk is quieter: an unmaintained, unpatched flag service sitting on your request path. Here is how to plan the move.

Read the post

· 9 min read

Build vs buy feature flags: when rolling your own actually makes sense

If you have five flags and one service, do not buy a platform. The interesting question is what breaks at flag fifty, and consistent bucketing is where it starts.

Read the post

· 9 min read

Feature flags vs feature branches: decouple deploy from release

Long-lived branches delay risk until the worst possible moment: a giant merge and an all-or-nothing release. Flags move that risk to a dial you control.

Read the post

· 10 min read

How to run a gradual rollout: 1% to 100% without incidents

A rollout plan is three decisions: the steps, the hold time at each step, and the metrics that stop the ladder. Here is a playbook you can copy.

Read the post

· 8 min read

Kill switches: recover from a bad release in seconds, not a redeploy

The cheapest incident is the one you end in seconds. A kill switch is the difference between a rollback war room and one click.

Read the post

· 9 min read

A/B testing with feature flags: measure impact before you commit

You already split traffic with every rollout. A/B testing is the same mechanism with a hypothesis attached, and it changes what you decide to build.

Read the post

· 9 min read

Open-source feature flags vs a managed platform: the real trade-off

The download is free; the pager is not. What self-hosting a flag system really costs, and the cases where it is still the right choice.

Read the post

Early access · launching soon

Ship your next release behind a flag

Join the early-access list and be first in line when spots open. Flat pricing, unlimited seats, no card required.

No card required · your data stays yours

Keep exploring

More feature flag resources