You can get a working version of Flux if you use the nightly builds of InfluxDB and Flux. There’s also a nightly of Chronograf that includes a builder interface. Those are linked at the top of the download page at https://influxdata.com/download.
Right now it’s under heavy development. The overall syntax of the language is mostly finalized, but there are many functions that aren’t yet implemented or other features that we need to add. There are also many situations where a query could cause the fluxd process to panic.
Our goal is to finalize the design of the language in the next two months and ship it with InfluxDB 1.7, which is slated for the end of August (or sometime in September if we slip). That version of Flux won’t have all the functions that we intend to implement, but it should be usable and useful for many different queries. We’ll have user facing documentation to go along with that release, like a getting started guide and API reference.
We’ll be shipping it later in the fall as an integrated part of the InfluxDB 2.0 alpha release. You can find all the Flux code here: https://github.com/influxdata/platform/tree/master/query. You can track issues related to Flux and their progress with the area query label in the platform repo. You can see the more technical Flux language spec here.
We welcome feedback, particularly at this stage of development. Feedback we get now can influence the design of the language before we lock things down!