Notes on feature flags, releases, and engineering from the Featureflip team.
A feature flag outage should be survivable: SDKs evaluate from cached values and fall back to your defaults. The costly failure is an SDK that never reconnects after the service returns.
Featureflip has an MCP server. Let Claude Code, Cursor, and other AI agents create flags, ramp rollouts, and clean up stale ones from your editor over the Management API.
Put your AI app's model, system prompt, and settings behind one feature flag: roll a new LLM out gradually with sticky per-user bucketing, and keep a fail-safe kill switch.
Featureflip now has a REST Management API. Create projects, flags, and targeting rules programmatically to seed flags in CI, sync environments, and build internal tooling.
A flag key is a record you query, not a label you read. The anatomy of a good feature flag naming convention, how to enforce it in CI, scale it by team size, and migrate off bad names.
A dark launch ships a feature to production with nobody seeing it; a soft rollout then reveals it to a few users at a time. How to run both safely with feature flags.
Add feature flags to a Python service: local in-memory evaluation, the pre-fork trap that silently freezes flags under Gunicorn and uWSGI, and how to initialize per worker.
Add feature flags to a Go service: a pure-Go SDK with no cgo and no dependencies, in-process evaluation with no per-request network call, goroutine-safe reads, and a clean shutdown.
Add feature flags to a Node.js backend: one shared client, local in-memory evaluation with no per-request network call, readiness gating, and a clean shutdown on SIGTERM.
Add feature flags to a Next.js app: evaluate on the server in the App Router or getServerSideProps to kill the hydration flash, then hydrate the client for live updates.
How to add feature flags to a React app with a provider and a useFeatureFlag hook that re-renders automatically, plus user targeting and tests without the network.
Feature flag flat pricing versus per-seat and usage-based: how each model scales as your team and traffic grow, with a real cross-vendor cost comparison.
A homegrown feature flag starts as one if-statement, then grows into targeting, SDKs, audit logs, and cleanup nobody owns. When to build, when to buy, and the true cost of each.
Progressive delivery is two ideas, not one: release progression and progressive delegation. Here's how feature flags turn a deploy into a dial you can turn up slowly.
Eight feature flag best practices for production teams — lifecycle tagging, flat decisions, both-state testing, progressive rollouts, governance, observability, and cleanup discipline.
Featureflip now supports first-class prerequisite flags — a child flag stays off until its parent serves the right variation, no application-side gating required.
Nine feature flag anti-patterns behind real outages — Knight Capital's $460M loss, Slack's day-long cascade, Google Cloud's June 2025 crash. Each with a fix.
Most feature flags are never removed, and one reused bit cost Knight Capital $460M. The four-step playbook to clean up flag debt safely.
Env vars configure where the process runs; flags decide what each request sees. Decision rule, comparison table, and 4 production gotchas inside.