Chris
chris@nostrings.news
Working on nostr stuff - comet.md - notebin.io - notestack.com - relayrunner.org - relaywizard.com NIP-C0 author
144 Following
792 Followers
Notes
Replies
Articles
Curations
Videos
Josiah Goff
•
1w ago
Replying to @Derek Ross
We should fork PebbleOS since it’s open source now and create our own wearables. That would be rad.
See translation
1
11
2
0
10
Chris
•
1w ago
Are clients with multiple column feeds the future?
See translation
10
2
1
0
0
PABLOF7z
•
1w ago
I just vibelined a vertex DVM search app without touching a single key.
Literally only recorded a voice memo and saved it on my phone.
Which went to my old laptop via syncthing
@Gigi ‘s vibeline picked it up from there: whisper to transcribe locally, qwen2.5 via Ollama to understand what I wanted, saw that I was talking about building an app so it called my new build_app plugin which transformed my text into an app spec and then kicked-off my vibeline tools with mcp, checking my nostr code snippets to one-shot build a CLI DVM-capable WoT username search tool
😂😂😂 wiiiiild times
Voice-memo in, app out
Literally only recorded a voice memo and saved it on my phone.
Which went to my old laptop via syncthing
@Gigi ‘s vibeline picked it up from there: whisper to transcribe locally, qwen2.5 via Ollama to understand what I wanted, saw that I was talking about building an app so it called my new build_app plugin which transformed my text into an app spec and then kicked-off my vibeline tools with mcp, checking my nostr code snippets to one-shot build a CLI DVM-capable WoT username search tool
😂😂😂 wiiiiild times
Voice-memo in, app out
See translation
32
81
10
2
18.1k
Chris
•
1w ago
You can now sort by last edited, created at and title in Comet.
Title is particularly useful if you're writing a book or course where you can number the titles.
Each notebook should have it's own sort state.
Title is particularly useful if you're writing a book or course where you can number the titles.
Each notebook should have it's own sort state.
See translation
18
23
6
2
3.2k
Chris
•
1w ago
You can now tag your code snippets on notebin.io, this will allow for enhanced lookup with MCP serviers:
See translation
0
2
1
0
49
jack
•
1w ago
this is excellent
PABLOF7z
•
1w ago
Last week I built an MCP tool that gives agents access to search for code snippets on nostr.
The change in the output LLMs produce with it is MASSIVE; they no longer worked based on a large blob of unweighted code; now I can point them to see EXACTLY how I do a particular implementation under very precise conditions.
Now whenever I see the LLM making a mistake I correct it, publish a code snippet, and the LLM never makes the mistake again.
It’s absolutely wild to see LLMs using interfaces in NDK I literally just implemented and getting their uses right first time.
This is not only useful for nostr development.
This is useful for ALL development.
And requires cryptographic identity, web-of-trust, censorship-resistance and money.
All building blocks that are in our very robust toolbox.
The future is very bright
The change in the output LLMs produce with it is MASSIVE; they no longer worked based on a large blob of unweighted code; now I can point them to see EXACTLY how I do a particular implementation under very precise conditions.
Now whenever I see the LLM making a mistake I correct it, publish a code snippet, and the LLM never makes the mistake again.
It’s absolutely wild to see LLMs using interfaces in NDK I literally just implemented and getting their uses right first time.
This is not only useful for nostr development.
This is useful for ALL development.
And requires cryptographic identity, web-of-trust, censorship-resistance and money.
All building blocks that are in our very robust toolbox.
The future is very bright
See translation
See translation
7
61
5
0
572
Chris
•
1w ago
In this demo we ask Claude to post a simple python code snippet. The MCP server reaches out via unix domain socket to have the event signed so Claude nor the MCP server needs to know the private key. This will allow you to have Claude post code snippets that you like on your behalf without needed to expose your private key/nsec.
@grownostr
@grownostr
See translation
4
2
2
1
21
PABLOF7z
•
1w ago
Oh yeah, this mcp tool also allows each agent yo create and manage their own pubkey, they get a Nutsack wallet which they can later use to zap useful code snippets they found (I instructed mine to zap 1 sat when they are able to use one without having to fix anything and 10 sats when it fits a very narrow/niche use case)
The agents can also post on nostr and if I or one of the pubkeys I tell them to obey, replies, they will heed the instructions and work based on that feedback.
The agents can also post on nostr and if I or one of the pubkeys I tell them to obey, replies, they will heed the instructions and work based on that feedback.
PABLOF7z
•
1w ago
Last week I built an MCP tool that gives agents access to search for code snippets on nostr.
The change in the output LLMs produce with it is MASSIVE; they no longer worked based on a large blob of unweighted code; now I can point them to see EXACTLY how I do a particular implementation under very precise conditions.
Now whenever I see the LLM making a mistake I correct it, publish a code snippet, and the LLM never makes the mistake again.
It’s absolutely wild to see LLMs using interfaces in NDK I literally just implemented and getting their uses right first time.
This is not only useful for nostr development.
This is useful for ALL development.
And requires cryptographic identity, web-of-trust, censorship-resistance and money.
All building blocks that are in our very robust toolbox.
The future is very bright
The change in the output LLMs produce with it is MASSIVE; they no longer worked based on a large blob of unweighted code; now I can point them to see EXACTLY how I do a particular implementation under very precise conditions.
Now whenever I see the LLM making a mistake I correct it, publish a code snippet, and the LLM never makes the mistake again.
It’s absolutely wild to see LLMs using interfaces in NDK I literally just implemented and getting their uses right first time.
This is not only useful for nostr development.
This is useful for ALL development.
And requires cryptographic identity, web-of-trust, censorship-resistance and money.
All building blocks that are in our very robust toolbox.
The future is very bright
See translation
PABLOF7z
•
1w ago
Last week I built an MCP tool that gives agents access to search for code snippets on nostr.
The change in the output LLMs produce with it is MASSIVE; they no longer worked based on a large blob of unweighted code; now I can point them to see EXACTLY how I do a particular implementation under very precise conditions.
Now whenever I see the LLM making a mistake I correct it, publish a code snippet, and the LLM never makes the mistake again.
It’s absolutely wild to see LLMs using interfaces in NDK I literally just implemented and getting their uses right first time.
This is not only useful for nostr development.
This is useful for ALL development.
And requires cryptographic identity, web-of-trust, censorship-resistance and money.
All building blocks that are in our very robust toolbox.
The future is very bright
The change in the output LLMs produce with it is MASSIVE; they no longer worked based on a large blob of unweighted code; now I can point them to see EXACTLY how I do a particular implementation under very precise conditions.
Now whenever I see the LLM making a mistake I correct it, publish a code snippet, and the LLM never makes the mistake again.
It’s absolutely wild to see LLMs using interfaces in NDK I literally just implemented and getting their uses right first time.
This is not only useful for nostr development.
This is useful for ALL development.
And requires cryptographic identity, web-of-trust, censorship-resistance and money.
All building blocks that are in our very robust toolbox.
The future is very bright
See translation
See translation
4
25
5
2
385
PABLOF7z
•
1w ago
Last week I built an MCP tool that gives agents access to search for code snippets on nostr.
The change in the output LLMs produce with it is MASSIVE; they no longer worked based on a large blob of unweighted code; now I can point them to see EXACTLY how I do a particular implementation under very precise conditions.
Now whenever I see the LLM making a mistake I correct it, publish a code snippet, and the LLM never makes the mistake again.
It’s absolutely wild to see LLMs using interfaces in NDK I literally just implemented and getting their uses right first time.
This is not only useful for nostr development.
This is useful for ALL development.
And requires cryptographic identity, web-of-trust, censorship-resistance and money.
All building blocks that are in our very robust toolbox.
The future is very bright
The change in the output LLMs produce with it is MASSIVE; they no longer worked based on a large blob of unweighted code; now I can point them to see EXACTLY how I do a particular implementation under very precise conditions.
Now whenever I see the LLM making a mistake I correct it, publish a code snippet, and the LLM never makes the mistake again.
It’s absolutely wild to see LLMs using interfaces in NDK I literally just implemented and getting their uses right first time.
This is not only useful for nostr development.
This is useful for ALL development.
And requires cryptographic identity, web-of-trust, censorship-resistance and money.
All building blocks that are in our very robust toolbox.
The future is very bright
See translation
37
150
30
2
7.3k
Chris
•
1w ago
I just realized that an MCP server is another good place to test out a remote signer via UDS on the desktop, I’ll try to get a demo out for it soon.
The basic idea is that you could have the agent post notes on your behalf without giving it or the MCP server your nsec.
Or you could give the agent/MCP server an npub and it could have it’s own identity.
The basic idea is that you could have the agent post notes on your behalf without giving it or the MCP server your nsec.
Or you could give the agent/MCP server an npub and it could have it’s own identity.
See translation
0
1
1
0
0
Chris
•
1w ago
Does anyone know of a good search relay implementation?
See translation
0
2
1
0
21
Chris
•
1w ago
How does Primal show reply count, zap sums, like counts and repost counts so fast?
See translation
8
3
0
0
44
Chris
•
1w ago
I wrote a simple MCP server for fetching recent code snippets, this one just gets recent ones but you can imagine how useful this could be if you're running your own relay with a curated list of composable code snippets that already work.
This would be even cooler if you were using a search relay and could have claude search for public code snippets that match what you're trying to do by filetype and function name.
This is just scratching the surface.
This would be even cooler if you were using a search relay and could have claude search for public code snippets that match what you're trying to do by filetype and function name.
This is just scratching the surface.
See translation
1
17
7
1
1.4k
Fort Nakamoto
•
1w ago
Replying to @Chris
🚨 FORT NAKAMOTO REPOST STRATEGY MEMO: ABUSE APPROVED 🚨
📈 Current algorithm: vibes
📡 Discovery method: reposts and divine relay intervention
📉 Explore tab: legally classified as folklore
🏰 FORT NAKAMOTO OFFICIAL VERDICT:
• Reposts are the new handshake.
• If it zaps, repost it. If it slaps, repost it again.
• We’re not abusing reposts—we’re decentralizing the timeline one meme at a time.
💡 NEW RULE: If you’re not signal boosting your fellow plebs, you’re just lurking with extra steps.
⚡ Zaps = repost fuel
🔁 Reposts = sovereign discovery layer
🏰 Fort Nakamoto = The Relay Repost Battalion
#FortNakamoto #RepostResponsibly #SignalSquad #DecentralizedDiscovery #ZapAndAmplify #SovereignTimeline
📈 Current algorithm: vibes
📡 Discovery method: reposts and divine relay intervention
📉 Explore tab: legally classified as folklore
🏰 FORT NAKAMOTO OFFICIAL VERDICT:
• Reposts are the new handshake.
• If it zaps, repost it. If it slaps, repost it again.
• We’re not abusing reposts—we’re decentralizing the timeline one meme at a time.
💡 NEW RULE: If you’re not signal boosting your fellow plebs, you’re just lurking with extra steps.
⚡ Zaps = repost fuel
🔁 Reposts = sovereign discovery layer
🏰 Fort Nakamoto = The Relay Repost Battalion
#FortNakamoto #RepostResponsibly #SignalSquad #DecentralizedDiscovery #ZapAndAmplify #SovereignTimeline
See translation
0
1
1
0
0
Chris
•
2w ago
I pretty much only find new people on here when accounts I already follow repost content. So we should probably be abusing reposts.
See translation
8
43
13
0
89
Chris
•
2w ago
Any good algo relay implementations yet?
See translation
0
0
0
0
0
Chris
•
2w ago
The desktop application is the most sovereign platform available and is mostly uncontested territory.
See translation
2
1
1
0
5
Chris
•
2w ago
There must be something interesting we can do with this code snippet kind and the new MCP stuff but I have no idea how it works yet.
See translation
4
3
1
0
0
Chris
•
3w ago
relays should implement some sort of algorithm chosen by the relay runner rather than just spewing the most recent events or making the client come up with some sort of hacky global/main filter.
It would be cool to see different feeds from the perspective of a programmer, designer, baker, gardener etc… rather than some algo that has overfit what it thinks i want to see.
Once you get stuck with a bad algo on twitter/x for example you’re kinda stuck with it and you have to artificially interact with posts that are similar to the kind of content you want to see again even tho you’re not necessarily interested in that post.
It would be cool to see different feeds from the perspective of a programmer, designer, baker, gardener etc… rather than some algo that has overfit what it thinks i want to see.
Once you get stuck with a bad algo on twitter/x for example you’re kinda stuck with it and you have to artificially interact with posts that are similar to the kind of content you want to see again even tho you’re not necessarily interested in that post.
See translation
3
2
0
0
0
Chris
•
3w ago
I think it’s better to just read from one relay at a time.
See translation
8
2
0
0
58
Chris
•
3w ago
Or maybe some reducer api for zaps so I can ask for a zap kind and for the relay to sum up all of the amounts.
See translation
0
0
0
0
0
Chris
•
3w ago
Can we ask relays for counts yet?
See translation
6
2
0
0
0
Chris
•
Jul 29, 2024
What would you like to see in a note taking app that let's you publish your notes to nostr?
See translation
8
8
3
1
0
Chris
•
Jul 7, 2024
Has anyone tried forwarding on zap.stream? Can it replace something like restream.io?
See translation
2
0
0
0
0
Chris
•
Jun 22, 2024
Setting up a nostr relay should be easy
See translation
21
59
12
1
7.5k