Nostr Web Client

GE, y'all.

I'm attempting to get MKStack working with a local LLM in order that I vibe code a NIP-72 Reddit-like application for #nostr, though I'm having a hard time dealing with it. LM Studio is treating me like garbage, I don't understand Ollama in terms of integrating it to Stack right now, and I'm having trouble finding an online LLM service that has a generous free tier that isn't Anthropic, OpenAI or any of these other big companies (DeepSeek isn't an option for me either, unless it's a locally downloaded LLM). If anyone has a recommendation for what LLM I should use, or how I could connect up Stacks locally in this manner I described without getting an improper JSON error in LM studio, that would be appreciated.

If local isn't an option, the above question for what online LLM I could use that has a generous free tier is what I #asknostr for today.

Reply to this note

Please Login to reply.

Discussion

SpiralCrunch 6mo ago

I am trying to do the same with a local LLM via Ollama (which works on it's own, but slow)

stacks clone mkstack -m devestral

stacks agent -m devstral

The above were given to me by Alex.

It has got me closer but still not able to make it work. It might be my hardware.

MK Fain 6mo ago

AFAIK none of this is really supported because we know these models won't provide a good experience - however if you succeed in hacking it let me know! In the long run we do hope to support more options.

PlebInstitute 6mo ago

Did set an API key when running stacks configure? Don’t leave it empty.

Derek Ross 6mo ago 💬 2

Technically, you have all of the context and MCP configs needed to use any LLM and any Agent. You don't need to use Stacks and it's AI provider options.

For example. Use Goose and Ollama.

PlebInstitute 6mo ago

I tried that with goose and LM Studio but working inside an MKStack Project fails. The same goes for stacks.

https://cdn.nostrcheck.me/0f6ad7e156bc6e410035cd6dff0ea69e15d22a1c85e84cad4b8ac5affe6459df/251f508686fb3326c7738659990fb6dbd82c3b2c3e59f0372096c78deecb0ffe.webp

PlebInstitute 6mo ago

Start your LM Studio server and set an API key via stacks configure. The value can be anything. Then do „stacks naddr1qv…“. Abort after the project was set up and enter the project directory. Them run „stacks agent -m “. Now you are one step further towards the next error 😄