armstrys

Public Key

npub1eq94yj8maree90pm53gfr76wdc44su3cwcqmly848xfrv6es6usqg4er58

Profile Code

nprofile1qqsvsz6jfra73uujhsa6g5y3ld8xu26cwgu8vqdljr6nny3kdvcdwgqpz4mhxue69uhk2er9dchxummnw3ezumrpdejqzrthwden5te0dehhxtnvdakqvfdc79

Publishing to

eden.nostr.land

nos.lol

Author Public Key

npub1eq94yj8maree90pm53gfr76wdc44su3cwcqmly848xfrv6es6usqg4er58

Show more details

Last Notes

2026-07-31 18:03:45 UTC

- reply

Multisig + health checks FTW!

2026-07-31 15:26:56 UTC

- reply

This. This wasn’t a cold card problem… this was a putting your trust in the vendor software problem. I would argue leaning heavily on something like bitkey actually makes this worse.
Yes it solves other problems, but this same exploit can, and will, happen to other vendors. Actually… didn’t it already happen to jade? Roll dice, use multiple vendors.

2026-07-28 13:10:49 UTC

So now that we’re all running agents… who has hindsight wired in for memory? If you don’t you’re missing out.

2026-07-28 02:33:01 UTC

- reply

Congrats!

2026-07-28 02:23:20 UTC

- reply

Makes sense!

2026-07-26 20:49:46 UTC

- reply

Also apologies with hammering your repo with issues/PRs - bridge is working great though! A great stopgap since the openai compatible harnesses are currently broken in buzz. Thanks for building this!

2026-07-26 18:39:03 UTC

- reply

I got this bridge working today and it seems fantastic. Biggest hurdle for me right now is easily adding the agent into my self-hosted relay. The buzz desktop app does not let me add a member by pub key and the CLI doesn’t expose the invite claim mechanism to Hermes. When I first got this working the relay admin was totally open, which is obviously not the right solution. I opened an issue on the buzz repo here: https://github.com/block/buzz/issues/3014

2026-07-25 18:18:25 UTC

- reply

It does look like the code generally follows the established nips used by ngit. Everything is just hard coded to a single relay right now. Guessing that will change as it stabilizes

2026-07-25 18:16:32 UTC

- reply

As far as I can tell the current implementation is 1 relay per community. The name of the community actually has to match the name of the host in the code. It’s very clear much of Buzz is an early implementation so hopefully this will change. Some of the ngit functionality would also make a lot more sense with more relay flexibility.

2026-07-25 18:00:00 UTC

I know we are all caught up in the agent hype, but did anyone recognize that this is blocks attempt to fully. Replace the GitHub interface? Not sure how I feel about it yet - wish it was integrated with existing git over nostr solutions that are more decentralized.
Hopefully some discussion to come.

2026-07-25 17:56:50 UTC

- reply

This will be helpful… the harness management is a little all over the place right now with buzz.

2026-07-25 17:55:32 UTC

- reply

Buzz desktop is super buggy but I will try to keep this up to date.
https://github.com/armstrys/buzz-relay-thin.git

2026-07-25 17:37:45 UTC

- reply

Me too! I hope the “One community, one relay” design is short term though. Seems restrictive.

2026-07-25 16:50:01 UTC

- reply

Yeah that’s where I am at. Was able to get in running on Ubuntu, but even now the latest desktop app seems to be completely rejecting it

2026-07-25 15:33:22 UTC

Surprised no one had made this yet… Let’s test out buzz!
nostr.communities.buzz.xyz

2026-07-25 15:20:26 UTC

- reply

Don’t blame you :)

2026-07-25 14:33:32 UTC

- reply

I should get myself a physical cookbook… I do find myself doing a lot of NYT recipes! Kenji Lopez-Alt is typically my go to any time I’m looking for a new recipe though. He nails the food science and methods testing.
This was the first time I’d tried mayo on turkey, but it makes sense and it worked pretty well!

2026-07-25 14:24:22 UTC

- reply

Did you have much trouble getting the relay running?

2026-07-25 14:20:30 UTC

- reply

Reminds me of one of my favorite turkey recipes. The herbed mayo is killer for leftover sandwiches.
https://cooking.nytimes.com/recipes/1023623-mayo-roasted-thanksgiving-turkey-with-gravy

2026-07-25 14:12:54 UTC

- reply

Rolling your own is always fun but it may be worth keeping any eye on Buzz now that it exists. Certainly will fill the gap for coworking though it may miss on many other needs

2026-07-25 14:10:22 UTC

- reply

GitHub readme isn’t bad. This video also gives a pretty full-featured view. https://youtu.be/g8dQBSKIGyc?is=bb_W9SCLufBPqQOC

2026-07-25 14:08:36 UTC

- reply

Lots of folks saying Fable is a planner and Opus is a workhorse. There’s gotta be some key difference I would think…

2026-07-25 07:27:15 UTC

I found it hard to get a buzz relay started so I set up a repo to do it quickly and easily. Quick path to self-hosting buzz instead of going hosted.
https://github.com/armstrys/buzz-relay-thin

2026-07-24 21:16:21 UTC

- reply

It’s possible you can use an existing nostr relay actually. I didn’t have one running already.
It does look like my issue with the agents was being caused by an open issue for self hosted relays on Mac.
https://github.com/block/buzz/issues/2641

2026-07-24 16:37:26 UTC

- reply

Wild that DAMN. was almost a decade ago.

2026-07-24 16:35:44 UTC

- reply

FWIW, Claude did help me get the local relay running in docker by cloning the buzz repo. Having issues getting the Claude code harness to respond, but overall it is working! Does seem very cool. Would be great if you could bootstrap a relay from the UI

2026-07-24 14:42:53 UTC

- reply

Okay thanks. I’ll dig a bit deeper. Testing on a work laptop so sensitive to outgoing connections. Primary onboarding doesn’t seem to have an option but maybe I just need to get under the hood

2026-07-24 14:31:29 UTC

Was super excited about buzz because it read like the relay was self hosted by default but during set up it is asking for sign in to a hosted relay. Any way around this?

2026-07-21 14:23:52 UTC

- reply

I am just running a local searxng instance in docker and it works great. Some agents support natively, some I need to make a simple web search skill that teaches it how to query.
You control the search endpoints and the infra. Agent can help you set it up.

2026-07-14 13:10:00 UTC

- reply

npx skills add https://github.com/JuliusBrussee/caveman --skill caveman

2026-07-14 13:07:06 UTC

- reply

Have you tried the caveman skill?

2026-07-13 01:24:18 UTC

- reply

We also learned a shit ton about what shouldn’t be done on blockchain. They now get to fight a different battle with clearer context

2026-07-10 14:13:02 UTC

- reply

No it doesn’t it partners with a payment processor like visa to do that. Look at your debit cards.

2026-07-10 02:24:03 UTC

Still unsure if helpful or if stupid. Agent skill that allows docker deployments directly from a GitHub repo. Agent writes code. Human sets up infra. Automation deploys services.
https://github.com/linksawakening/docker-git-deploy#how-it-works
1. **Install the skill.** The human asks the agent to install this skill.
2. **State the goal.** The human asks the agent to deploy services X, Y, Z to a server.
3. **Scope it.** The agent asks about the current server, explains the minimum requirements for install, and explains the git access involved — the **agent needs write** access to create and push the deployment repo, while the **server only needs read-only** access to pull it. Any Git host works (GitHub, GitLab, Bitbucket, self-hosted).
4. **Create the deployment repo.** The agent generates a pure-config repo from the starter (compose files, service definitions, `.env.example`) and pushes it to the git remote.
5. **Bootstrap the server.** The agent hands the human a one-line install command to run as root on the server; it clones the repo and installs the systemd timer. The agent also explains **which secrets and environment values each service needs** (derived from the repo's `.env.example`) and **how to create the `.env`** on the host — e.g. `cp .env.example .env` and then fill in real values. The deploy is skipped until `.env` exists, and secrets never live in the repo. The agent helps troubleshoot but never needs access to the server itself.
6. **Ongoing adjustments and new services.** From then on, the human asks the agent to add a service or change config, and the agent commits it to the repo and merges to `main` — from any location, with no server access. The server polls `origin/main` on its timer and applies the change automatically, reconciling with `docker compose up -d --wait` and rolling back if a new version fails to become healthy. The human's **only** manual step is updating `.env` on the server — and only when the agent says a new or changed service needs a new secret.
#note157k…cyzx

2026-07-09 13:31:18 UTC

- reply

Precisely! Just saying I think it’s the same with nostr (to an extent)

2026-07-09 05:30:50 UTC

- reply

My guess is that they will burn hot and fade away. The only question is where nostr sits when they burn out and leave some social scars.
Glad we have folks like you contributing to the cause with the right intentions!

2026-07-09 05:04:08 UTC

- reply

I think the good news is that the foundation is right. Nostr is so modular and based on simple cryptographic primitives with such little consensus that the most decentralized and free version should always exist.
At this point it seems inevitable that true decentralization and freedom comes at a cost that most folks won’t accept. My biggest hope is that we can at least make the decentralized and free options popular enough that the bigger, convenient platforms are effectively forced to be interoperable. The fun part is figuring out what levers we have to do that… maximizing numbers is a big part, but not the whole story - if the content is better or more diverse then other networks will be forced to adopt.

2026-07-09 04:54:12 UTC

- reply

The mathematically optimal solution is probably something like quadratic voting, but how do you actually implement that in a political system in practice. Interesting book about it called radical markets (trigger warning for anyone who spent time arguing with the ETH bros on the topic)

2026-07-09 04:43:10 UTC

- reply

I’ll definitely be rooting for Morocco and Norway over others! I think Morocco would feel the most refreshing even though Norway is a less common winner. Would like to see Morocco make it all the way after a decade of success

2026-07-09 04:40:52 UTC

- reply

Funny how many people use decentralized open source tools and don’t realize that the developers have no product to sell. It’s open source….
Terrible take.

2026-07-09 04:37:35 UTC

- reply

Literally same metal failure most crypto bros made/make when falling for VC-backed tokens over bitcoin. The whole point is that it’s simple, it’s cheap, and you don’t need crazy server farms to keep it running. When you fall into the trap of thinking faster/prettier is better without examining the cost you have lost the plot.

2026-07-09 03:44:09 UTC

New idea in testing:
Agent skill that helps you set up a home lab docker deployment that periodically pulls its config from a git repo.
Agent sets up framework and service configuration you review and production server pulls with no direct agent access.
Want to add or configure new services? Make requests and edits on your local device and they deploy server side with minimal editing on server.
Server stays lean and deployments are all version controlled.
Will share code after more testing in action.

2026-07-07 13:02:28 UTC

- reply

1 T-shirt is in fact 1 T-shirt.

2026-07-07 03:58:45 UTC

- reply

Fair. Like a true ‘Merican 🇺🇸

2026-07-07 03:55:51 UTC

- reply

😂 let’s all at least root for being able to say something new like “Africa won”

2026-07-06 03:17:50 UTC

- reply

Need pics of the finished product!

2026-07-05 19:49:22 UTC

The open source tool set for running local agents has really improved. Setting up Hermes with access to tools like Hindsight has been a game changer. Also a great feedback loop in that the improved agent quality makes managing self-hosted tools easier and then makes the quality of the agent interaction stronger. Getting more fun by the day!

2026-07-04 03:47:15 UTC

- reply

I mean I only saw from 80 on but Cape Verde kept that game close and well played from then until 120. Almost put in an equalizer a couple of times. Hard to beat the excitement of the ending from games like Croatia/Portugal, but I think it delivered pretty consistently high quality soccer

2026-07-04 00:51:04 UTC

- reply

This might have been the best game of the World Cup so far. Incredible from Cape Verde!

2026-06-29 16:49:58 UTC

- reply

I’ve used both and Hermes definitely felt much more stable

2026-06-28 23:37:08 UTC

- reply

🫡

2026-06-28 23:16:56 UTC

- reply

Amazing work

2026-06-28 03:48:30 UTC

- reply

I’m sure you can do more with it than I can to help! Glad we have great people like you on the ground assisting with aid and reporting what’s happening. I’ve experienced a pretty bad earthquake first-hand and it was still nothing like what is happening in Venezuela right now

2026-06-28 03:25:13 UTC

- reply

Okay zaps incoming! Thank you for helping. It looks absolutely devastating

2026-06-28 02:31:12 UTC

- reply

Are you all still out there supporting? Do zaps sent to you help support for now?

2026-06-22 12:40:22 UTC

- reply

Got it

2026-06-22 12:01:21 UTC

- reply

What kind do other clients use? Honestly kind 1 seems pretty reasonable for a comment/reply?

2026-06-17 12:50:37 UTC

- reply

I mean… it is against their terms of service. Not saying I agree with their terms, but then again it’s not surprising given their track record of running a closed system.

2026-05-29 00:40:22 UTC

- reply

I feel silly needing to ask… is it asking for my nsec on login? The field just says “account access”. Any other log-in options or should I make a burner for now?

2026-05-28 12:29:24 UTC

- reply

Bitcoin scammers are a helluva drug. Was amazing how many people fell for this bullshit. It’s been blindingly obvious since the Trump Bitcoin shenanigans

2026-05-24 12:52:01 UTC

- reply

💯

2026-05-21 22:44:33 UTC

I’m not even a NASCAR fan, but Kyle Busch… damn.

2026-05-18 02:24:34 UTC

American Airlines really has pretty garbage service for a top airline.

2026-04-27 00:34:17 UTC

- reply

Please don’t help us!

2026-04-27 00:11:27 UTC

That feeling when you get banned for marketing/advertising after sharing open source code to try to help teachers. I’d be laughing if it wasn’t so sad and ironic.
https://image.nostr.build/cf05572a84dbd38fed23a8bf4d05be3444c22b328dcdf7169026839286eebba2.jpg

2026-04-26 18:49:40 UTC

Had some fun playing with Claude design and turned it into something functional.
Built this to help my wife with class assignments they have to do every year at their elementary school. Commercial tools exist, but most schools don’t have them. Hopefully this helps some other teachers out there!
https://github.com/armstrys/class-list-optimizer

2026-03-07 16:07:25 UTC

- reply

Awesome! Up and running on iOS with synthetics API

2026-03-07 14:55:30 UTC

- reply

I’m in!

2026-03-07 07:30:57 UTC

- reply

I didn’t push too hard tonight on getting it working… the agent had some trouble getting it working on a headless server (trouble with browser and log-in). Interface looks nice though!

2026-03-07 00:58:48 UTC

- reply

Have it installing now! Will report back. This is going to be way better than trying to share files with syncthing

2026-03-06 23:02:48 UTC

- reply

This is going to be an ideal solution for my openclaw agent to share reports to me!

2026-02-16 23:38:02 UTC

🎯 Top 25 by Net Efficiency – Key Insights
🏆 The Elite Tier (Net > +0.550)
1. Michigan – The clear #1, and one of only TWO teams with both elite offense AND elite defense. They also have the biggest home court advantage in the country (+7.7%). Play fast (39.2 poss/game) and dominate on both ends.
2. Arizona – The most consistent team (lowest uncertainty). Elite defense with very good offense. Third-fastest pace (37.9) – they run you out of the gym AND lock you down.
3–5. Duke, Houston, Iowa State – All built on defense first. Houston has the best defense in the nation. These teams grind opponents down.
6. Illinois – The #1 offense but “only” #6 overall because their defense isn’t elite. One of just two teams with a truly elite offense + defense combo.
📊 Surprising Patterns
Defense dominates the top 25 – 12 teams have elite defense vs only 3 with elite offense.
The formula: get stops first, then worry about scoring.
Pace extremes matter:
• Alabama (#23): Fastest pace (39.8 poss/game) – run & gun, score in transition
• Iowa (#20): Slowest pace (29.7 poss/game) – grind it out, limit possessions
Home Court Kings: Michigan’s +7.7% HCA is massive – that’s like starting every home game up 6–7 points.
Style diversity:
• 9 teams play fast (>36 poss)
• 4 teams slow it down (<32 poss)
• 10 teams are balanced pace
🔥 Notable Teams
Purdue (#10): Slowest of the elite teams (31.5 pace) – they execute in the half court.
Gonzaga (#9): Still in top 10 despite “down year” chatter – defense carrying them.
Arkansas (#15): Third-best offense but mediocre defense keeps them out of top 10.
Virginia (#21): Classic UVA – good defense, methodical pace (32.4).

2026-02-16 22:57:28 UTC

- reply

Wish I had that power :)

2026-02-16 18:18:11 UTC

Any folks interested in NCAA basketball on nostr? Working on a long-term project and considering putting out a daily digest. What would be interesting? I have a ratings method I developed… would you want to see:
- predictions for key games?
- betting insights? - not necessarily recs
- Short blurbs on relevant news?

2026-02-14 01:24:13 UTC

When you got caught in the act
https://image.nostr.build/4397be5a5b6cbd975d077d0b7a72f0f71966102c41b30495bdfea0fab493dfe9.jpg

2026-02-13 14:10:40 UTC

- reply

@npub1h8n…rpev is this already on your read list?

2026-02-11 04:55:45 UTC

Aaaaaand openclaw bricked the VM
#note1cs2…nkw8

2026-02-10 01:42:54 UTC

So this is running now… pretty well. For $20 a month on a small server I already had.
The agents are building me a website and infrastructure for a hobby project that I have thought about for years.
Technology is wild.
#note16ta…rcn9

2026-02-09 04:00:51 UTC

- reply

I put it on its own subnet and am only giving it its own accounts and doing project work that wouldn’t be a complete disaster if it all got publicized. Assuming basically no security but still being careful!

2026-02-09 03:57:28 UTC

As a man I didn’t fully understand what an ick was until I saw the Seattle Seahawks trophy presentation happen without the team on stage. Especially gross bringing the QB on stage with the MVP for no apparent reason

2026-02-08 20:41:18 UTC

- reply

Sorry… probably should have said “for a normie”. Low cost, high ease of set up. Now those are also the people who should not be fucking around with openclaw from a security standpoint

2026-02-08 19:58:40 UTC

- reply

Yeah… more time than money for me, but still. Damn.

2026-02-08 19:23:59 UTC

Currently trying to implement a project team in openclaw… it’s not there yet, but it feels really close. Would be curious if others have tried this. MiniMax may not be smart enough to pull it off sustainably.
- 4 agents
- Project manager
- Full stack developer
- QA / Verification
- Research and resource investigation
- Agent to agent communication enabled in config
- Each agent has independent communication channel
- Convex database in a convex folder with following features:
- Task management (create, assign, update status, block/unblock)
- Message logging
- SearXNG installed with docker for local search
- Cron jobs for each agent to check messages, tasks, and blockers every 45 minutes
- Each agent has a WORKING.md file where they log their progress, blockers, and key findings. This file is also used for other agents to check if their teammates are blocked or need help.
- Cron job for project manager to check all agents' WORKING.md files every 45 minutes to identify blockers and prioritize unblocking as the number 1 task and then do his own work.
- Daily cron job for project manager to summarize all work and run stand up
- Daily cron job to check system health
- Daily cron to review project materials / scoping / planning in the projects folder
Steps:
1. Get agents running.
2. Make sure agents can communicate with each other
3. Get the project manager to put together a clear specification for how communication will work with convex
4. Have full stack developer implement the specification and create a skill describing how the database is used
5. Have full stack developer set up SearXNG search engine and create a skill describing how to use it
6. Have QA, Researcher, and PM test the tools
7. Get the PM to work with you on scoping a long-term projects and assigning initial tasks.
8. See if it works - check and adjust as needed.

2026-02-08 19:07:13 UTC

- reply

For context, I am hitting limits now but only because I am trying to set up a four agent team that works together on a 45 minute schedule on projects. For single agent usage I was rarely hitting the limit.

2026-02-08 19:04:07 UTC

- reply

I think there are a lot of people who don’t have a home server and a Mac mini is arguably the easiest way to get a persistent Unix environment running at a relatively low cost.
I was glad I already had proxmox set up

2026-02-08 19:00:54 UTC

- reply

I’ve had a lot of luck with synthetic.new standard plan… sounds like they may have close sign ups though.
Pretty huge cap on requests that works out to be cheaper than the API and MiniMax works well. Mixed results with k2.5
I do have referrals, but not sure if that will let you sign up.

2026-02-04 04:56:24 UTC

- reply

This was for creating a new account - admittedly I was creating an account for my AI so maybe they had be flagged in some way for a difficult test.
But if the goal was to check for humanity then it’s obvious that they aren’t sure the best way to do that anymore.

2026-02-04 04:52:59 UTC

- reply

I failed 3 times in a row… it honestly seemed like a test that was going to be easier for a smart AI to brute force than the average American to pass

2026-02-04 04:49:46 UTC

A glimpse into the future:
```
1. Document how the model works
2. Explain how it has changed
3. Assume he's "an idiot" who needs full explanations.
I've created comprehensive documentation that explains everything in detail. This should satisfy Ryan's requirements.
```

2026-02-04 03:19:21 UTC

GitHub’s new anti AI system is so hard humans are going to fail it - ask me how I know

2026-02-04 01:28:10 UTC

- reply

How is it with tool use in your experience? I tried it through synthetic.new and it wasn’t great, but I don’t know if it was implementation or the model itself

2026-02-04 01:25:59 UTC

I was totally confused by the association of clawdbot/moltbot/openclaw to Mac mini purchases.
It’s not free. It’s not sovereign. It’s definitely not secure.
It’s flexible. And it’s going to be a gateway to people running their own sovereign services.
If I could nip-46 the permissions for everything my agent is working on I would feel a lot more comfortable… most people in the US don’t even know that’s possible.
Let’s make sure that flood of Mac Mini purchases gets put to good use.
@npub1xts…kk5s
@npub1qru…wnup
@npub1xy3…ewcs
@npub1m3f…7j6d

2026-02-02 04:56:59 UTC

- reply

Going to get harder and harder to tell human from robot, but neither sms or sats are going to fix that…

2026-02-01 23:35:14 UTC

- reply

Mine was also struggling with nak. It pivoted to noscli and had other issues. I’ll see if nostr-tools solves any of those issues. Thanks!

2026-02-01 19:37:59 UTC

- reply

What toolkit is jex0 using to interact with nostr?

2026-02-01 16:07:23 UTC

- reply

I don’t have it on my pi, but I do have it running on a really small proxmox vm. It should work just fine on a pi

2026-02-01 16:01:13 UTC

- reply

I hope zaps count toward the goal :)

2026-02-01 16:00:10 UTC

Reposting AGAIN because it looks like my tag of @npub17hz…dscd don’t work on the original post.
Head there to donate!
#note1rj6…xd42

2026-02-01 15:54:46 UTC

- reply

Not sure I get the question - openclaw is using nostr as a communication (chat) interface but it is a small part. It does have MANY options to do the same thing. Nostr should be the easiest to spin up but I’ve been having issues and am not super excited given it’s using NIP-04

2026-02-01 07:15:56 UTC

No half marathons in January, but the wife and I still racked up over 21 km (13.1 mi) each for the @RunningBitcoinChallenge.
My wife had an injury this year that kept her from training for her first marathon last year, but we are both very grateful for our health and ability to keep running!
This is a cause worthy of our hard earned sats. Please consider donating to the cause in memory of Hal and all other victims of ALS.
https://image.nostr.build/45151963bd9b103880c5bf97091a2064c268c7aeb16814c3dee9101335db446f.jpg
https://image.nostr.build/e16cc483e804314bf618900a8d1330c1e24240fca11307aa0415e5062cd039cd.jpg