indefractus
bent

don’t yuck someone else’s yum

Public Key

npub1vgvrulklmv9v9ff8vl2hk64etv9eyjca0a7yekrse85fgrtyf8dqlxv0sc

Profile Code

nprofile1qqsxyxp70m0akzkz55nk04tmd2u4kzujfvwh7lzvmpcvn6y5p4jynkspzemhxue69uhhyetvv9ujuurjd9kkzmpwdejhgqg5waehxw309aex2mrp0yhxgctdw4eju6t0qyxhwumn8ghj7mn0wvhxcmmv4dgdj4

Publishing to

relay.primal.net

relay.damus.io

nos.lol

Author Public Key

npub1vgvrulklmv9v9ff8vl2hk64etv9eyjca0a7yekrse85fgrtyf8dqlxv0sc

Show more details

Last Notes

2026-05-28 03:39:46 CEST

- reply

So wait, sanity check: how the fuck would Tony Robbins know with such certainty that we are going to have AGI in 3 yrs? Is he an AI researcher? Is he running a frontier lab? Does he have a glass jar filled with Dario Amodei’s farts that he huffs on the reg? Oh btw I think his $40/month chatbot is a great value. Its tagline is “The power of Tony’s strategies - in your pocket.” Yeah, that mofo is in a lot of peoples’ pockets alright.
Now I need to go meditate about why I have such animosity towards Tony Fucking Robbins (TFR) since I never gave him one second of thought even after I watched his amazing performance in Shallow Hal 25 yrs ago.

2026-04-20 05:10:26 CEST

It was tough to get a decent functioning memory setup for my openclaw personal assistant. A lot of trial & error, going back to the drawing board, nuking different setups. But I think I finally have landed on something that works for me.
The key was the Honcho (https://app.honcho.dev/) memory plugin for OpenClaw. I’ve used it enough at this point to be able to say that it makes a significant difference over OpenClaw’s out-of-the-box memory tool. I used to cuss out my agent every five minutes because it was so stupid and forgetful between sessions, now it’s actually a helpful personal assistant and fun to work with.
I now have a four-layer, fully automated memory setup:
1. Honcho is the base layer - ambient memory that absorbs every chat turn. It runs on my @nprofile…mjrt in a Docker container next to the OpenClaw container. Honcho is backed by a PostgreSQL database that constantly distills, summarizes, reasons, and dreams about the chat logs. It provides all the rich history and context to the LLM.
2. Agent operational memory in the form of MEMORY.md and YYYY-MM-DD.md files in the OpenClaw agent workspace. The agent maintains these automatically. A shell script prunes operational memory after 30 days to keep things tidy.
3. My Obsidian vault lives on my phone but is synced to Umbrel with Syncthing so it also accept writes from the agent. It functions as my 2nd brain. A nightly agent cron job reviews operational memory and automatically promotes stuff worth keeping into my 2nd brain.
4. The agent pushes to GitHub its project files, scripts, and anything else it produces for me. Fully synced so it just pushes files to the private repo.
I set the whole thing up from my iPhone. Linux systemd runs a shell script that does Honcho Postgres dumps nightly to Umbrel’s Documents Folder, then Umbrel backs everything up to an external drive. I’m not paying for Honcho’s cloud service. Just paying for OpenAI tokens to run the OpenClaw model and Anthropic tokens to run the Honcho in the background.
Next up is to give my agent a Claude Code terminal and a Google Workspace and see what I can make it do.

2026-03-20 23:44:11 CET

- reply

No. It’s an intergalactic holiday due to Chuck Norris having served as our planetary. Ambassador to MoM-z14, which is a mere13.5B light years away. An easily-traversable distance for Chuck Norris.

2026-03-13 18:16:31 CET

- reply

Was just reading Chapter 1 on my Kindle, although I should be working 😉. Impressively gripping, 47 to go!

2026-03-13 04:08:30 CET

- reply

I don’t know if this is answering your question, but I used ChatGPT to help me build and deploy OpenClaw in a custom Docker container on my Umbrel Home device. In other words, I’m not using the official Umbrel OpenClaw image from the Umbrel App Store. I’m using the official one from the OpenClaw GitHub site. I ssh into the container and it just works awesome. I talk to the agent through my browser and over Signal and it’s working good so far. I have it sitting behind https and Tailscale for security and remote access when I’m not on my home LAN. It has a Google Workspaces account and a Brave browser API key. And I’m slowly trying to build some multi-agent capabilities. This setup is after probably six weeks of frustrating trial and error because I’m not a pro or anything. But ChatGPT has been a huge help.

2026-03-13 03:22:46 CET

- reply

Well then it’s decided. You’re a Pythagorean. Is that you on the left? 😆
https://blossom.primal.net/98387691cff59729db8c8f3b6ff8629691b8d36a2bf54207faa42255a01b4e96.png

2026-03-13 03:08:27 CET

- reply

If anything think I called you an ancient philosopher. Pythagoras believed everything was made of numbers. He believed in the perfection of the number 3 as it represents past, present, and future. He also had some weird aversion to beans. How do you feel about beans?

2026-03-13 02:31:17 CET

- reply

Oh, so you’re a Pythagorean?

2026-03-11 19:45:21 CET

- reply

Definitely not yay (rejoice) but also not yea (affirmative). I’m anti-trough except when there’s no other option. And usually there isn’t. A trough is one step up from a latrine, but just barely. So I’d say troughs are a necessary evil. However, I might do the trough over a Porto-pottie, which makes Porto-Potties pure unadulterated evil. Not just nay, but hell naw. Now maybe there is such a thing as a high quality porcelain trough experience in some airport in Dubai or a high end bar in Manhattan. But I’ve never seen one. That picture is probably the best I’ve seen. Ever.

2026-03-11 19:28:01 CET

- reply

Damn, dude, is that seriously the number? 4k? I had no idea. That’s minuscule.

2026-03-11 08:48:17 CET

- reply

Still skeptical. Every IRS form and worksheet is numbered and published online. The OP says it accompanies the the1099-DA. No mention of this supposed worksheet in the 10-page instructions.

2026-03-11 01:40:27 CET

- reply

Pardon my skepticism. They “just” created that form? It looks like my grandma got creative with a spreadsheet then made a photocopy of a photocopy. Link to the gov publication pls? All I could find are a couple of shady crypto “ news” sites that mention this story but don’t link to any official info.

2026-02-25 12:25:07 CET

- reply

Elite honesty tier 😆

2026-02-15 15:53:12 CET

- reply

I should also say that part of my motivation is to be able to cancel my ChatGPT Plus subscription and just use the API. But another notable, perhaps not-surprising thing: one full day of me chatting with OpenClaw using gpt-5.2 costed about $20 USD. Might cost half that after I dialed back thinking and unnecessary tool calls, but it still was more expensive on a daily basis than the Plus subscription. For some of those calls I probably didn’t need to use OpenClaw, and asking ChatGPT would have been fine and would have saved money.
For comparison, the API cost of gpt-5.2 is $14 per million tokens out; 5-mini is $2; Nano is $0.40. I switched to mini this morning and it seems totally capable for the kinds of conversation I’m having at this stage, i.e., I’m not calling for deep thought and leading-edge domain analysis. I’m just having it reference its own context, self-configure, and teach me how to use it. So I would not recommend connecting OpenClaw to an expensive model initially. Wait until you’re ready to give it the kind of work that warrants a premium cost.

2026-02-15 05:44:21 CET

- reply

I spent today configuring it by sitting here chatting on my iPhone on my private Tailnet. I use an iPhone app called GoCo Editor to remote into the Umbrel so I can see exactly how OpenClaw is handling its/my files. Using that approach I got it to
- create folders containing project plans it wrote for me,
- update its short term and long term memories,
- reorganize files and delete unwanted ones,
- set up cron jobs to send me reminders weekly, monthly, and daily based on my instructions
- write Python scripts and cron jobs to prune its file system weekly and monthly,
- create indexes to help me locate files in the future
- advise me when to switch to a less expensive chat model
-rewrite its code to reduce the number of tokens I was burning on inefficient tool calls
- plan for building additional agents that will all live in separate project-based group chats with me on Signal
I have been surprised by its strategy for managing memories. As we DM back and forth, it creates a daily memory file (yyyy-mm-dd.md) and stores whatever important memories we both decide should go in that file during the day. That’s like my daily curated journal that keeps growing with a new one each day. The agent’s basic instructions, which are fully modifiable if I wanted to ratchet them up, tell it not to build a dossier, to respect boundaries, and to earn trust through competence by being “good.” I can fully audit its memories at any time. It keeps the past few days’ memories easily accessible in its context window for continuity’s sake. Then it keeps a separate long term memory file (MEMORY.md) for things I specifically want it to keep in context for longer periods. Beyond that I toggled on the feature to have it create a sqlite database of vector embeddings from all my historical chat transcripts. It can reference that db to recall the meanings - not exact content - of its entire history of conversations with me. And of course I can ask it to read in any file from outside its immediate context and db to (re)/introduce any specific info I want. All of this can practically stay local on Umbrel so I’m not feeding a bunch of context to a third party. But I am hitting OpenAI’s API with my prompts plus a vectorized payload. A stretch goal would be to move the LLM API calls to a local server running Ollama or LM Studio so that nothing leaves my home network. This assumes that some basic personal assistant functions don’t require a multi-trillion parameter frontier model running in the cloud.

2026-02-14 07:48:59 CET

- reply

I know ppl are springing for Mac Minis, but I have an Umbrel Home. It runs 24x7 so the agent is always on yet isolated from my main computer. I went to the App Store, clicked to install OpenClaw and let it walk me through setup. Connected it to my fav LLM (needed an API key). Connected it to Tailscale for a free private encrypted network. Dug up an old iPhone and got a Google voice line (free) so I can DM my bot on its own Signal app channel. Boom, I’m fully mobile and chatting with my agent. It can modify its own code, so I accomplished quite a bit by just talking with it over Signal. I’m not terribly tech savvy, but I’m slowly building its personal assistant capabilities, and I gotta say it’s pretty useful so far.

2026-02-13 04:00:35 CET

- reply

I’ve been entranced by OC for the past couple weeks. It’s running in my Umbrel Home device and connected to Signal over Tailscale so I can take it anywhere on my phone. I’m a project manager, not a coder, and the fact that it can modify its own code has gotten me out of so many jams whenever I do something stupid and break it. Cron jobs were busted, I chatted with it for a couple hours, and now they’re locked in. I just tell it what’s wrong, maybe paste some log output. Then it thinks for a couple minutes and fixes itself. And the more I use it the smarter it gets. So amazing.

2026-02-05 05:20:46 CET

- reply

The openclaw doc page seems like it would be helpful. https://docs.openclaw.ai/channels/telegram#telegram
Assuming you can figure out how to get the token from BotFather on Telegram, the OpenClaw part I think would involve going into Config, clicking the “Raw” button to expose the openclaw.json file, editing the channel section so it reads just like the minimal config section in the OpenClaw docs. Then click save and click update.
The other way to edit the json file is to go into the OpenClaw shell through the Umbrel GUI, Settings -> Terminal -> OpenClaw Terminal. Brew install nano if you don’t have it. Then use nano to edit the file. Then go back to the GUI, right click OpenClaw, and restart it. Your Telegram settings should load.

2026-02-05 04:31:55 CET

- reply

Sorry Im not a telegram user. I’m going to try Signal. I have found that the bot is helpful, so if you just ask it it will tell you how to configure it, and it will even modify its own openclaw.json file or any other files in its workspace. So you don’t need to struggle through the config section. gpt-5.2 was getting expensive ($14 per million tokens) so I switched to 5.1-nano $.40 per million tokens and the bot still gives helpful guidance for setup purposes.

2026-02-04 01:07:07 CET

- reply

I spent way too many hours over the past three days trying to get OpenClaw installed on Umbrel using various configurations of Portainer with Caddy, then tried nginx, and Tailscale. I was able to connect to the UI but it wouldn’t pair with the URL token. I was feeling totally defeated.
Then I checked the App Store and saw the app was there. I downloaded it, dropped in my OpenAI API key, and it connected no problem. But the bot wouldn’t respond. The issue (per troubleshooting log) showed it trying to connect to Anthropic not OpenAI. So I had to go into config under Models and manually enter openai/gpt-4o. Now it’s working. Thanks for getting the app out so quickly! 👍

2026-01-30 02:46:24 CET

- reply

Remember when everyone thought Darius Rucker’s name was Hootie? That was some buuullsheeot.

2026-01-28 01:47:06 CET

- reply

to chew bubble gum and kick ass…

2026-01-26 13:31:21 CET

- reply

Mongols and British and Russians, oh my

2026-01-17 01:49:46 CET

- reply

Glorious feast!

2026-01-17 01:47:40 CET

- reply

Oh, I see. I hadn’t heard about the Celsius scandal but just read up on it. As a fairly new Bitcoiner I keep hearing all these marketing pitches about borrowing against one’s BTC. I wondered how it works and knew there had to be a catch. As I now understand, there are better and worse ways to do it. A better way is to keep the LTV sufficiently low so you don’t get wrecked it the collateral value drops precipitously; ensure transparent margin call and liquidation rules, use a smart contract or escrow agent, don’t give away ownership/don’t allow the lender to rehypothecate. My guess is that companies marketing these deals are structuring them in ways that transfer risk to the borrower so the company can make greater profits. If I ever have enough BTC to justify leveraging it, I’ll be careful who I do business with. Thanks for recommending this as a learning opportunity.

2026-01-16 01:28:08 CET

- reply

I tried to navigate the terms and conditions associated with a Fold account and eventually got overwhelmed. The basic Ts & Cs are 33 pages not incl Privacy Policy. To link my bank account to the Debit card would require agreeing to the Plaid Privacy Policy (34 pages) and also Astra Ts & Cs (13 pages). Then if I want Exchange Services I have to agree to BitGo’s Ts & Cs. So my customer info would just proliferate to all these companies. It seems complicated. Plus, as I understand it, the only way to not pay fees when moving my money onto the Debit card is to set up direct deposit from an employer. So I’d need to really commit to moving away from my bank to take advantage of that. Otherwise it’s a1.5% fee to move money onto the account (capped at $25 per transaction) plus another 1.5% for BTC buys unless I pay $100 annually for the Fold+ program to avoid the exchange fee. I can see the potential benefit, but it seems like I’d need to redirect virtually all my net salary, all my bill payments, and all my credit card spending through that account and/or their credit card to really come out ahead. That’s a total upending and switchover away from my bank checking account and current credit cards. I understand that’s exactly what Fold is angling for. Not saying I like my bank. I actually dislike my bank and the paltry interest they pay on my balances. Are the benefits of switching worth it? If I went all in to rearrange my current setup the rewards could stack nicely. So far, by simply downloading the app, I have accumulated 75 free sats from daily spins of the wheel. But I guess I need to have 30,000 sats before I can withdraw to an external wallet. Oh well, that spinner pays better than some nostr zaps!

2026-01-15 23:52:51 CET

- reply

What happened? I’ve never studied the risk(s)

2026-01-14 00:59:14 CET

- reply

Dang those are high numbers for UK. Proportional to the US population, that would be like 61,000 arrested here.

2026-01-14 00:20:59 CET

- reply

I listened to the interview. Keonne said his BitCoin funds were frozen, so pulling together money for his own defense became a problem because he earned his salary in BTC. I think he pleaded guilty to lesser charges in exchange for dropping three more serious ones. But yeah, after hearing that interview I read the 51 page sentencing memorandum. Many screenshots and direct quotes from texts showing his own words. I assume he had no idea feds were monitoring or would dig that stuff up or else he and his associate wouldn’t have said things exactly the way they did.
I saw there was a grand jury indictment. So a group of ppl heard some set of purported facts and decided to indict. Obviously prosecutors frame it one way to achieve their desired ends, and the guy who says he’s been unjustly treated frames it the other way in public.
If you’re a free speech absolutist, if you think the dark web should be unfettered, or if you lean anti-govt, you might side with him and be sympathetic about signing the petition. Or you might think he’s just a developer who wrote some code that someone unfortunately didn’t want to exist.

2025-12-31 23:28:02 CET

https://blossom.primal.net/98287b4eaa4528de41e1cfd45cf4494453e9191b7f4fa2979b6e7bef8edc6571.jpg

2025-12-20 16:31:19 CET

- reply

Hopefully I justified my existence for today with the aid of Shakespeare and sixty minutes of free time. Behold my 8-bit platformer Nostrland Nomads. The ostrich and the HODLer (little green man with a cauldron) team up to stack BTCs. DOGE coins fall from the sky and the ostrich can fly around and zap them into BTCs with his laser beam eyes. If a DOGE coin hits the ostrich, he lays an egg that sprouts a mushroom. If the ostrich eats the shroom the sky turns psychedelic and the platforms start shifting around in space. I think the goal is for the HODLer to stack as many BTCs before...something happens. I haven't figured that part out yet 😄
https://blossom.primal.net/f610bc9d65669aa68d29a8a3c1b8b2ecbf8aafcc6f76688c3882794b11430b9c.png

2025-12-20 04:00:23 CET

- reply

No worries. Use hash tag asknostr if you have questions.

2025-12-18 14:01:01 CET

- reply

Well that one lifted my mood 😂

2025-12-18 13:40:29 CET

- reply

Not an Australian here, so asking out of curiosity: how does the govt discern between bitcoin sitting in a wallet as an investment vs sitting in a wallet for spending? Is the differentiator whether or not it was purchased on an exchange?

2025-12-14 19:44:24 CET

So, yeah, apparently part of wrapping my head around nostr will have to include understanding bot behavior. Hence the cat pictures bot thinks I want to zap it. Nah.
My initial assumption was that these bots are using natural language processing for targeting. ChatGPT thinks “…sometimes yes, but usually it’s much dumber than that.”
Further, it says:
When a bot follows you, probability-wise:
~70–80%: dumb follow-spray
~15–25%: keyword trigger
~5–10%: graph-based
<1%: real NLP targeting
So the m.o. for follow-spray purportedly is
Subscribe to recent kind:1 notes on 5–20 popular relays.
Extract pubkeys from:
• Anyone who posts
• Anyone who replies
• Auto-follow them
And the m.o. for keyword-trigger purportedly is
Subscribe to kind:1 notes with keyword filters
• Simple string matching or regex
• Follow authors whose posts contain keywords

2025-12-14 19:10:21 CET

🤔 So wait… was it the earlier post referencing my cats, or was it my mention of AARP that caused the astrology bot to glom on to my account? Drats…pigeon-holed again.

2025-12-14 18:13:37 CET

- reply

Thanks! I had a similar thought when I read Primal’s Ts & Cs, but at the same time I understand how they’re not locking-in ppl in the way my mind is sometimes wont to presume. They have a right to control their servers and capitalize on the market opportunity for premium services etc. I think their privacy policy seems fair, honest, and less restrictive than what one gets from many other corps in daily life.
One premium service is a larger amount of content storage. So this is making me realize that if I want to ensure my content (e.g. photo or video upload) endures without being evicted at some future point then I would need to either upgrade to theirs or some other hosting service with a content distribution network (AWS or whatever). Fair enough. As I understand it, the note events (absent the attached file content) will persist on the nostr relay network as long as I publish to multiple relays for redundancy.

2025-12-14 17:01:52 CET

To wit:
Me: In one paragraph, what would Sartre say about nostr?
ChatGPT (5.2): Sartre would likely see nostr as a vivid experiment in radical freedom and responsibility: by stripping away centralized authorities and anchoring identity in self-generated keys, it throws the individual back onto themselves, forcing each user to confront the fact that meaning, truth, and community are not guaranteed by institutions but must be actively created and sustained through choice and action. He would admire its refusal of “bad faith” in the form of algorithmic curation and corporate mediation, while warning that nostr does not absolve anyone of responsibility—on the contrary, the absence of a governing center intensifies it, because every post, relay choice, and social norm becomes an existential commitment for which no platform, moderator, or system can be blamed.

2025-12-14 16:59:26 CET

After signing up to nostr it took me a minute to grasp how awesome it is that my account is *mine* 🤯 and I can choose whatever client I want. I have been trying to suss out which one is right for me. I feel drawn to @nprofile…rnrm , which I think looks and feels great on iOS and web, but damn I haven’t been able to rule out some of the others yet. I think they each have decent pluses and less apparent minuses. I notice a fair amount of overlapping functionality across them and it seems to be just the implementation or emphasis on certain protocol features that slightly differs from one to the next. I’d be interested in knowing from more experienced users whether there is a real differentiator (e.g. wallet functionality, search/filter configurability, etc.) or whether I should consider this a largely aesthetic decision. ChatGPT would probably make me a feature comparison matrix, but seems like overkill - I dunno.
I understand that the decision is also, to an extent, use case driven, and that not all users have the same use case. The other thing I think probably may matter is that some of these offer an upgrade plan with access to more features. I imagine that could probably be a factor at some point. Finally, it could also be true that the need for “decision” is an illusion because any so-called choice is completely reversible in this context of absolute freedom. Condemned to freedom on nostr. Very Sartrean.
Thoughts?
#asknostr
https://blossom.primal.net/3d754bcdf6949d8d5a6ac858e292abb120fdf3d23975039b6f28cf1a5ddff3c3.png
https://blossom.primal.net/75acca3a370d450dfc1e593bf3c326b320814c198e996a5bc6fe841d89723eac.png

2025-12-14 15:32:15 CET

GM nostr.
The cats usually wake me up, but my wife took feeding duty and I ended up sleeping in a bit. I awoke with a startle from a vivid and highly unusual dream. Yikes, I hope that shit never happens irl.
Came downstairs and was greeted by the sonic onslaught of AC/DC’s Back in Black blaring from her phone - a video with Amish ppl and sleds strapped to a horse & buggy spinning out in a parking lot. This led to a morning coffee convo about whether AC/DC are still touring (yes in 2025), and another similar convo about The Rolling Stones (yes in 2024). The tour was sponsored by AARP. Seems un-fucking-real but apparently true.
Further research revealed that two days ago WSJ published an article called “The wisdom of Keith Richard’s at 82.” It contains this line: “An octogenarian, Mick can still be seen dancing on the ashes of Western Civilization.” Oh, the imagery.
Enjoy your day 👍

2025-12-14 03:36:34 CET

👉🏻 #noob here. I deleted FB and Twitter years ago for reasons that folks here can probably understand. And lately I have felt like YT is sucking my soul out through my eyeballs.
Almost by chance I caught @nprofile…04ye's bitcoin after dark episode. While he was riffin’ and just speaking his truth he said “For those of you who aren't yet on Nostr, you probably should be…”
I said to myself WTF is nostr? Honestly, in that moment something clicked. I drove home and immediately started figuring it out….
I’m glad I did because this thing is the shiz and I hope it keeps growing with unstoppable force.
I still have a lot to learn about this movement and this tech, but I’m happy to be here. I have found some really helpful resources here and elsewhere online. In January I’m going to get an @nprofile…r599 set up with @nprofile…mxeg Hub and the journey will continue.
From THE Bitcoin Podcast: DO THE F***ING MATH | BITCOIN AFTER DARK EP 1 (BAD001), Dec 6, 2025
https://podcasts.apple.com/us/podcast/the-bitcoin-podcast/id1694392423?i=1000739951153&r=224
This material may be protected by copyright.