Edit: I see you, making edits to the readme to make it sound more human-written since I commented ;) https://github.com/gavrielc/nanoclaw/commit/40d41542d2f335a0...
A hundred times this. It's fine until it isn't. And jacking these Claws into shared conversation spaces is quite literally pushing the afterburners to max on simonw's lethal trifecta. A lot of people are going to get burned hard by this. Every blackhat is eyes-on this right now - we're literally giving a drunk robot the keys to everything.
- Created its own github account, then proceeded to get itself banned (I have no idea what it did, all it said was it created some new repos and opened issues, clearly it must've done a bit more than that to get banned)
- Signed up for a Gmail account using a pay as you go sim in an old android handset connected with ADB for sms reading, and again proceeded to get itself banned by hammering the crap out of the docs api
- Used approx $2k worth of Kimi tokens (Thankfully temporarily free on opencode) in the space of approx 48hrs.
Unless you can budget $1k a week, this thing is next to useless. Once these free offers end on models a lot of people will stop using it, it's obscene how many tokens it burns through, like monumentally stupid. A simple single request is over 250k chars every single time. That's not sustainable.
> Skills over features. Contributors shouldn't add features (e.g. support for Telegram) to the codebase. Instead, they contribute claude code skills like /add-telegram that transform your fork.
I’m interested to see how this model pans out. I can see benefits (don’t carry complexity you don’t need) and costs (how do I audit the generated code?).
But it seems pretty clear that things will move in this direction in ‘26 with all the vibe coding that folks are enjoying.
I do wonder if the end state is more like a very rich library of composable high-order abstractions, with Skills for how to use them - rather than raw skills with instructions for how to lossily reconstruct those things.
Apple containers have been great especially that each of them maps 1:1 to a dedicated lightweight VM. Except for a bug or two that appeared in the early releases, things seem to be working out well. I believe not a lot of projects are leveraging it.
A general code execution sandbox for AI code or otherwise that used Apple containers is https://github.com/instavm/coderunner It can be hooked to Claude code and others.
I think most people fail to estimate the real threat that malicious prompts can cause because it is not that common, its like when credit cards were launched, cc fraud and the various ways it could be perpetrated followed not soon after. The real threats aren’t visible yet but rest assured there are actors working to take advantage and many unfortunate examples will be seen before general awareness and precaution will prevail….
Last time I checked, having a continuously running background process considered as a daemon. Using SQLite as back-end for storing the jobs also doesn't make it queueless.
/nit
Thankfully the official Agent SDK Quickstart guide says that you can: https://platform.claude.com/docs/en/agent-sdk/quickstart
In particular, this bit:
"After installing Claude Code onto your machine, run claude in your terminal and follow the prompts to authenticate. The SDK will use this authentication automatically."
Minor nitpick, it looks like about 2500 lines of typescript (I am on a mobile device, so my LOC estimate may be off). Also, Apple container looks really interesting.
https://maordayanofficial.medium.com/the-sovereign-ai-securi...
At least 42,665 instances are publicly exposed on the internet, with 5,194 instances actively verified as vulnerable through systematic scanning.. The narrative that “running AI locally = security and privacy” is significantly undermined when 93% of deployments are critically vulnerable. Users may lose faith in self-hosted alternatives.. Governments and regulators already scrutinizing AI may use this incident to justify restrictions on self-hosted AI agents, citing security externalities.This project violates Claude Code's Terms of Service by automating Claude to create an unattended chatbot service that responds to third-party messaging platforms (WhatsApp, and what you add ...).
The exact issues:
1. Automated, unattended usage - The system runs as a background service (launchd) that automatically responds to WhatsApp
messages without human intervention (src/index.ts:549-574)
2. Building a bot service - This creates a persistent bot that monitors messages and responds automatically, which violates restrictions on building derivative services on top of Claude
3. Third-party platform integration - Connecting Claude to WhatsApp (or other messaging platforms) to create an automated
assistant service isn't an authorized use case.
The README itself reveals awareness of this issue at line 41:
**No ToS gray areas.** Because it uses Claude Agent SDK natively with no hacks or workarounds, using your subscription with your auth token is completely legitimate (I think). No risk of being shut down for terms of service violations
(I am not a lawyer).
The defensive tone ("I think", "I am not a lawyer") indicates uncertainty about legitimacy. While using your own credentials doesn't automatically make automated bot services compliant—Anthropic's TOS restricts using their products to build automated chatbot services, regardless of authentication method.
The core violation: transforming Claude Code into an automated bot service that operates without human intervention, which is explicitly prohibited.I’m confused as to what these claw agents actually offer.
I assume this is to keep the footprint minimal on a Mac Mini without the overhead of the Docker VM, but does this limit the agent's ability to run standard Linux tooling? Or are you relying on the AI to just figure out the BSD/macOS equivalents of standard commands?
Quick Start
git clone https://github.com/anthropics/nanoclaw.git
Is this an official Anthropic project? Because that repo doesn't exist.Or is this just so hastily thrown together that the Quick Start is a hallucination?
That's not a facetious question, given this project's declared raison d'etre is security and the subtle implication that OpenClaw is an insecure unreviewed pile of slop.
Openclaw is very useful, but like you I share the sentiment of it being terrifying, even before you introduce the social network aspect.
My Mac mini is currently literally switched off for this very reason.
My gut reaction says that I don't like it, but it is such an interesting idea to think about.
1. You can live in the future, and be at the bleeding edge of the latest AI tech, reaping the benefits. Be part of the solution.
2. You can stay in the past and get left behind, at the mercy of those who took the risks.
I realize you used Claude Agent SDK on purpose but I'd really like to this to be agent agnostic. Maybe I'll figure that out...