Everything I publish, in the order I published it: dispatches from Twitter, videos from YouTube and notes from my vault, archived here in full — text, images, video and links — so they outlive the platforms.
922 dispatches since January 2012 · page 11 of 37
Post
2 replies · 3 likes · 435 views
~$1.68 per day. That's what having a nanoclaw-based assistant costs me.
Prediction: In less than 90 days, all channels that we thought were safe from spam & automation will be so flooded that they will no longer be usable in any functional sense: iMessage, phone calls, Gmail.
Oof. I immediately went to my VPS and checked because I couldn't remember exactly if I ended up on the correct github repo (and hence cloned the right thing) Good news:
My single most important learning after installing, customizing and working with @openclaw and @Gavriel_Cohen's nanoclaw:
The biggest risk was hiding in plain sight. And the only way to discover it was by actually exploring claws.
In order to make my point, I first need to briefly walk you through some of the "known unknowns" risks that I had on my radar:
Fear 1: That claws would be simply useless and I would have to face the social shame of trying something that really was just a joke. (Turns out, it IS actually very engaging, useful and even fun.)
Fear 2: Having credentials such as API keys and auth profiles on the server (to make claws actually useful) and then getting those credentials stolen and abused. (Could still happen, but also easily managed with spending limits.)
Fear 3: Customizing my assistant with code and skills from the internet which would lead to any of the above and more subtle attacks such as semantic worms, anti-memetic ███████, cognitive root-kits etc. (Could and probably will happen but we'll deal with it when that is the case)
However. ALL of my fears were DWARVED by the biggest, unanticipated, unmonitored and unmanaged blind spot:
THE FUCKING UTTERLY INSANE TOKEN BURN AND POTENTIAL LOSS OF MY ANTHROPIC MAX SUBSCRIPTION!
In hindsight, it literally sat there right smack in the middle of all the other fears and risks.
And at the same time, I did not have this on my radar AT ALL.
And of course: Once I was aware of it (Thank you @RealistSec for your help), I could start managing it.
My personal solution is that I've given my assistant the ability to sleep, dream and forget. Meaning, at night, a cron-job takes the session data from the previous day, checks for a few important bits to remember and then clears out all session storage. In the morning, the agent starts out fresh and with a free mind.
But my solution isn't what matters here. What matters to me is that I was reminded how action produces information. And that the things that actually hurt often are not what we anticipated. And that the peeps at @AnthropicAI stayed cool (thank you) while a few of their servers went up in flames.
I briefly believed that "cron job" or "claws" would be the words of the year 2026.
Now I believe it'll be "AINXIETY".
...
"Don't believe the hype!" said the great prophet Carlton Douglas Ridenhour and his friend, the wise Bodhisattva William Jonathan Drayton Jr., in unisono, in 1988.
And they were right. Do not believe the hype. If this AI doomerism is getting to you: Take a break. Take care of yourself. Take a nice long shower. Eat something yummy. Call a friend!
And yes, also "do look up." There is so much borderline magic stuff happening right now. You ARE sitting front-row at Dylan's 1965 performance at the Newport Folk Festival.
Things change. They always have and will. The good news is: We humans are actually really good at change. YOU are really good at this.
Looking back, I am so grateful and lucky that my godmother Diane had subscribed to @NatGeo throughout the 80s. You could pick any issue from the stack and travel the world instantly.
It's crazy: I can still open any issue from 1985 and recognize every cover, most of the photos and ads and even some of the stories.
Of course, Jame Gurney's Rockwell-y soybean tableau was, and still is, an all-time favorite!
So, on behalf that 8 year-old kid growing up in West Germany and this one adult who has lived in many countries since:
Thank you Diane and thank you National Geographic.
❤️
OMG, I only discovered this now: You can also peruse those foldable supplements in stunning detail. Oof.
Looking back, I am so grateful and lucky that my godmother Diane had subscribed to @NatGeo throughout the 80s. You could pick any issue from the stack and travel the world instantly.
It's crazy: I can still open any issue from 1985 and recognize every cover, most of the photos and ads and even some of the stories.
Of course, Jame Gurney's Rockwell-y soybean tableau was, and still is, an all-time favorite!
So, on behalf that 8 year-old kid growing up in West Germany and this one adult who has lived in many countries since:
Thank you Diane and thank you National Geographic.
❤️
In the past week, I've had lunch with, respectively, 5 friends of mine. All of them top-notch non-technical founder-types with solid track-records.
Each conversation quickly circled in on AI claws and how they're experimenting with it, including the development and operations part. I LOVE that they're finally enabled to do this on their own and I love seeing that creative/hacker-y spark in their eyes. It feels like they've waited ages for a technical sparring-partner that could deliver at the thought-of-speed.
I am so happy for them and some of their self-made products are impressive. Technically and product-wise.
What I also noticed, and that's a surprise for me, is that when they go into very deep detail on their firewall configuration, I feel sad because I feel they've been kinda nerd-sniped/FOMO'd into this rabbit-hole and I am not sure how and when they'll find their way out again.
PS: Hand-written by Alexis, delivered to Twitter by Alexis' nanoclaw assistant
[continued thought]
So, this is NOT about me wanting to gatekeep vibe-coding and vibe-shipping. I absolutely welcome it. In fact, they leveled up so quickly that our technical conversations as real as they get.
But they also look like they haven't slept in days, they're pale. They seem more neurotic than usual. They're a bit less charismatic and a bit more lost. And I guess that makes me feel as if I'm "losing the good ones" to the infinite techno/AI rabbit-hole that I learned to maange and navigate a long time ago.
PPS: Here's how I used Telegram/nanoclaw to post this (and thus avoiding getting sucked into the maelstrom of my Twitter feed) - That is using the desktop/Chrome skill I created. That's running in a permanent VM alongside the assistant.
The editorial policy I've set in the skill is that nanoclaw is only there to help format and deliver.
In the past week, I've had lunch with, respectively, 5 friends of mine. All of them top-notch non-technical founder-types with solid track-records.
Each conversation quickly circled in on AI claws and how they're experimenting with it, including the development and operations part. I LOVE that they're finally enabled to do this on their own and I love seeing that creative/hacker-y spark in their eyes. It feels like they've waited ages for a technical sparring-partner that could deliver at the thought-of-speed.
I am so happy for them and some of their self-made products are impressive. Technically and product-wise.
What I also noticed, and that's a surprise for me, is that when they go into very deep detail on their firewall configuration, I feel sad because I feel they've been kinda nerd-sniped/FOMO'd into this rabbit-hole and I am not sure how and when they'll find their way out again.
PS: Hand-written by Alexis, delivered to Twitter by Alexis' nanoclaw assistant
PPS: Here's how I used Telegram/nanoclaw to post this (and thus avoiding getting sucked into the maelstrom of my Twitter feed) - That is using the desktop/Chrome skill I created. That's running in a permanent VM alongside the assistant.
The editorial policy I've set in the skill is that nanoclaw is only there to help format and deliver.
In the past week, I've had lunch with, respectively, 5 friends of mine. All of them top-notch non-technical founder-types with solid track-records.
Each conversation quickly circled in on AI claws and how they're experimenting with it, including the development and operations part. I LOVE that they're finally enabled to do this on their own and I love seeing that creative/hacker-y spark in their eyes. It feels like they've waited ages for a technical sparring-partner that could deliver at the thought-of-speed.
I am so happy for them and some of their self-made products are impressive. Technically and product-wise.
What I also noticed, and that's a surprise for me, is that when they go into very deep detail on their firewall configuration, I feel sad because I feel they've been kinda nerd-sniped/FOMO'd into this rabbit-hole and I am not sure how and when they'll find their way out again.
PS: Hand-written by Alexis, delivered to Twitter by Alexis' nanoclaw assistant
Alright - just keeping a tab on getting a nanoclaw assistant to a point where I feel like it's okay to stop fiddling with the setup:
Yesterday, I had precisely 125€ in Anthropic API credits. The basic nanoclaw /setup and /customize yesterday came in at around ~15€.
Then, running /setup-telegram, testing out agent swarms and building a simple status-dashboard that I can access from my laptop added another 20€.
So, I've spent ~35€ so far and still feel like I'm getting fantastic value.
And, what I LOVE about nanoclaw is the clear separation between using Claude Code for setting things up, customizing them, running ops etc. vs. using the Anthropic API for running the actual assistant(s).
Currently, I'm using Anthropic's Agent API for both, Claude Code and the assistants.
Given how I'm using Claude Code to build/setup/ops the system, I feel it would be fair use to use my Pro subscription here instead of paying directly for API tokens.
And given how the Pro subscription (as someone on HN wrote a few days ago) is, after all, like a ticket for one person to an all-you-can-eat buffet, I feel it would not be cool to use my Pro subscription here and it's totally right to pay for API tokens.
So, these are some of my thoughts here from today.
Next up: Since agents are run in ephemeral containers (yay!), I'm currently having Claude Code build me a separate docker container that runs a persistent desktop environment (KDE FTW) that will include a Chrome browser that both, I (via SSH/VNC) and the agent (via CDP) will have access to and can use.
Aaaaand 10€ worth of Anthropic API tokens later, here we are!
What I do like is having a clear distinction between "This is the Claude Code conversation to build this" and "This is the WhatsApp conversation to use this"
Yup. Knowing when I'm working with the "systems-architect" and when I'm interacting with the "assistant" really helped me decide where to fine-tune what.
With OpenClaw I wasn't sure (and that's on me still) where that boundary was. I had the TUI chat, the main web chat, the messenger chat. And I wound up using all of them for configuration as well as assistance.
Alright - just keeping a tab on getting a nanoclaw assistant to a point where I feel like it's okay to stop fiddling with the setup:
Yesterday, I had precisely 125€ in Anthropic API credits. The basic nanoclaw /setup and /customize yesterday came in at around ~15€.
Then, running /setup-telegram, testing out agent swarms and building a simple status-dashboard that I can access from my laptop added another 20€.
So, I've spent ~35€ so far and still feel like I'm getting fantastic value.
And, what I LOVE about nanoclaw is the clear separation between using Claude Code for setting things up, customizing them, running ops etc. vs. using the Anthropic API for running the actual assistant(s).
Currently, I'm using Anthropic's Agent API for both, Claude Code and the assistants.
Given how I'm using Claude Code to build/setup/ops the system, I feel it would be fair use to use my Pro subscription here instead of paying directly for API tokens.
And given how the Pro subscription (as someone on HN wrote a few days ago) is, after all, like a ticket for one person to an all-you-can-eat buffet, I feel it would not be cool to use my Pro subscription here and it's totally right to pay for API tokens.
So, these are some of my thoughts here from today.
Next up: Since agents are run in ephemeral containers (yay!), I'm currently having Claude Code build me a separate docker container that runs a persistent desktop environment (KDE FTW) that will include a Chrome browser that both, I (via SSH/VNC) and the agent (via CDP) will have access to and can use.
So far, using Anthropic's new agent teams has been working well. Technically speaking.
Spawning teams was the impetus for building the dashboard I showed earlier. I almost immediately had a sense of "I don't know what's going on here. I need visibility."
So now at least I see my main agent using TeamCreate and Task delegation (See the bottom-half of screenshot for example)
...which may go up to 14€ soon by way of this statement: t.co/xa38J8cbnB
Note that I fully support @Hetzner_Online here because, as a on-and-off customer since 21 years, their service has been absolutely flawless and always worth every cent. They care so I'm totally happy to pay whatever they need to make it work for them, too.
Add to that about 4€ for creating a new mentor-like "role" that I want openclaw to "play" and setting it up with a daily cron-job.
So, ~11€ in total in Claude Code API credits.
Honestly, this feels right. I actually now think that a lot more onboarding flows should be chat-based and Claude Code was, again, just a solid partner. Walked me through the steps one-by-one, fixed problems quickly.
Pretty impressed.
Alright - just keeping a tab on getting a nanoclaw assistant to a point where I feel like it's okay to stop fiddling with the setup:
Yesterday, I had precisely 125€ in Anthropic API credits. The basic nanoclaw /setup and /customize yesterday came in at around ~15€.
Then, running /setup-telegram, testing out agent swarms and building a simple status-dashboard that I can access from my laptop added another 20€.
So, I've spent ~35€ so far and still feel like I'm getting fantastic value.
And, what I LOVE about nanoclaw is the clear separation between using Claude Code for setting things up, customizing them, running ops etc. vs. using the Anthropic API for running the actual assistant(s).
Currently, I'm using Anthropic's Agent API for both, Claude Code and the assistants.
Given how I'm using Claude Code to build/setup/ops the system, I feel it would be fair use to use my Pro subscription here instead of paying directly for API tokens.
And given how the Pro subscription (as someone on HN wrote a few days ago) is, after all, like a ticket for one person to an all-you-can-eat buffet, I feel it would not be cool to use my Pro subscription here and it's totally right to pay for API tokens.
So, these are some of my thoughts here from today.
Next up: Since agents are run in ephemeral containers (yay!), I'm currently having Claude Code build me a separate docker container that runs a persistent desktop environment (KDE FTW) that will include a Chrome browser that both, I (via SSH/VNC) and the agent (via CDP) will have access to and can use.
@SpringStreetNYC@openclaw@virustotal@good_BOT_boi Oh! And if you haven't updated for a while, the old version also used to use the main session UUID + it's history to spin up every 'heartbeat' check in every 30mins - so if you had a huge session, that would consume all your tokens on a check-in too,
GitHubGitHub - nanocoai/nanoclaw: A lightweight alternative to OpenClaw that runs in containers for security. Connects to WhatsApp, Telegram, Slack, Discord, Gmail and other messaging apps,, has memory, scheduled jobs, and runs directly on Anthropic's Agents SDKA lightweight alternative to OpenClaw that runs in containers for security. Connects to WhatsApp, Telegram, Slack, Discord, Gmail and other messaging apps,, has memory, scheduled jobs, and runs dir...↗ github.com
Claude Code DocsOrchestrate teams of Claude Code sessions - Claude Code DocsCoordinate multiple Claude Code instances working together as a team, with shared tasks, inter-agent messaging, and centralized management.↗ code.claude.com
