🚀 Feature Request: Examples of using KV with local dev

[paddymul]

Describe the solution

I am trying to develop a workers app. I wish there were examples of using the kv commands in a tutorial

My project.toml looks like this

kv_namespaces = [
    { binding = "FLATBUSH3",id = "24de8aba6ca3470e98741e5d2ecb59d7", preview_id = "6dd38aed6466411ba3cbaad37930ac35" }
]

I am trying to load a JSON file into the kv cache. I have tried the following commands:

wrangler kv:bulk put  --namespace-id 6dd38aed6466411ba3cbaad37930ac35  --preview  ../kv_bulk.json  
wrangler kv:bulk put --binding FLATBUSH3 --preview  ../kv_bulk.json      

I can see the keys via wrangler kv:key get and wrangler kv:key list. My code doesn't see the keys that I have added.

Further, it is possible to store to the KV in JS/TS code as json or an arraybuffer. There is no documented way to do this via the wrangler kv tools. https://developers.cloudflare.com/workers/runtime-apis/kv/#types

I would expect worked examples showing each one of these use cases.

[AdityaSher]

+1 my namespace is null on dev but has data in prod (both namespaces have data on the cf dashboard)

[vassbence]

Hi!

It seems like in v3 there is no way to access/edit the local KV data from the CLI. I think adding a --local flag similar to how D1 operates would solve this.

As for my use-case: I use KV in my application as a runtime-configuration store. I'd need to put a few rows into KV locally to run my e2e tests with full code coverage.

[penalosa]

https://developers.cloudflare.com/kv/reference/kv-commands/ is the documentation for KV commands locally. In Wrangler v4, KV commands will interact with local data stores by default, and will interact with remote data if you add the --remote flag.