wagle

Feature flags · remote config

Ship behind flags,
evaluated at the edge.

Wagle is a multi-project, multi-environment feature-flag system where every flag lives in memory and evaluates locally, with no database on the hot path, built for very-high-traffic apps.

Take the desk See pricing

xxhash64 · bucket % 10000 · in-memoryWhy & how it works, including where it's less strong →

Why wagle

A control desk for every rollout.

No database on the hot path

Every flag lives in an in-memory, (project, environment)-partitioned index. Eval is a hash and a table lookup, no round-trip, ever.

Multi-project, multi-environment

Flags are keyed by (project, environment) day one. Same flag set across dev, staging, prod, with a different rollout in each, promoted explicitly.

Deterministic bucketing

xxHash64 over {project}:{env}:{flag}:{entity}, mod 10000. The same entity always lands in the same bucket, so a 5% rollout is a stable 5%.

Targeting, segments & rules

Compose rules over segments and attributes with an ordered fall-through to a default variation. Typed values, not just booleans.

Versioned config history

Every write bumps a per-project version and appends to history, attributed to an actor. Read a whole version torn-free, never a mixed snapshot.

SDK sync, edge-ready

SDKs pull one initial snapshot and evaluate locally. Server-only flags stay server-side. Built for very-high-traffic apps.

The desk, live

What you actually get.

Untouched screenshots of the console working a real project: one bool flag, one string flag with a rolling split, one JSON config with a rule and a rollout, and the history those edits leave behind.

The flag list showing three typed flags with variation counts and rollout bars — 1 · The desk One project in production: three typed flags and a segment. Everything the rest of this tour touches is on this screen.

The flag editor on a simple boolean flag, enabled in production — 2 · Start simple Start simple: a boolean kill-switch, enabled in production, serving its default.

The segment editor showing the beta-testers segment with its conditions — 3 · Define who Define the audience once: a segment is a reusable predicate, one level deep on purpose. Beta testers, spelled out.

Close-up of a rollout rule splitting traffic 60/25/15 across three variations — 4 · Split the rest Beta testers get the invite; everyone else rolls 60/25/15. Deterministic per entity: the same id always lands in the same slot.

The flag editor on a JSON flag with four variations, a targeting rule and a rollout — 5 · Compose Then compose the full thing: a JSON config with four variations, an IF rule, and a rollout fallthrough. Typed, not stringly.

The Evaluate tool showing an entity's deterministic bucket and the lit serving path in the targeting graph — 6 · Verify Verify exactly who gets what: type an entity, bucket 9194 of 10000, the lit path shows what serves and why. Same id, same bucket, every SDK.

The flag history timeline showing versioned, attributed edits with path-level JSON diffs — 7 · Audit Everything above left a trail: every write versioned and attributed, JSON edits diffed down to the inner key.

The flag editor with the dev panel open on the MCP tab, showing the agent tool call for the current flag — 8 · Automate Your AI agents work the same desk: the dev panel shows the MCP tool call for the flag on screen. Agents are revocable principals, not shared keys.

The SDK demo resolving a flag locally, showing reason, bucket, snapshot version and a local no-network source — 9 · Ship it And this is how it reaches your app: the real SDK live in the page, seeded once, kept fresh on its poll, every eval local with no network hop. Server-only flags never leave the server.

Pricing

You pay for syncs, nothing else.

Evaluations run locally and are always free: the only call that reaches wagle is an SDK sync (the config pull). So the bill scales with one number: how many times your SDKs sync.

price ∝ syncs · evals free

Starter

5k syncs included, then pay as you grow.

$5per month

Start free trial

5k SDK syncs / month
then $1 per additional 1k
Unlimited local evaluations
Unlimited flags & segments
Unlimited projects & environments
30-day config history
Email support

Team

More volume, cheaper overage.

$49per month