On GitHub turn the artefact's URL into a permalink by pressing `y`. This automatically changes the URL in the browser.
10.03.2026 13:09
👍 0
🔁 0
💬 0
📌 0
On GitHub turn the artefact's URL into a permalink by pressing `y`. This automatically changes the URL in the browser.
Yes, there could be. I have read non Claude Code use distorted their usage insight.
Lock-in , sorry missed the dash.
Stopped my Claude Code subscription because they do not allow using their plan outside of Claude Code. Which is well in their right to do but I consider it to be a lock-in move.
(I use Opencode as my main coding agent.)
Lock-in below refers to all, currently AI, tooling being geared towards Github.
'Physically' moving is dead-easy, when developing locally only remote is different.
`tea` can be used a alternative to to GitHub CLI.
Also: sometimes, I feel like I am making it too hard on myself.
(E.g. caring about privacy and big tech, using Jujutsu over Git, trying to avoid agent lock in not using Claude Code ATM.)
Started to use Codeberg.
A European-based, privacy-friendly, non-profit alternative to Github.
It’s only when you move you realize how locked-in you are.
(E.g. LLMs often assuming Git or GitHub.)
docs.codeberg.org/getting-star...
codeberg.org/hanlho
We are aware of recent reports regarding targeted phishing attacks that have resulted in account takeovers of some Signal users, including government officials and journalists. We take this very seriously. 1/7
Well ... that does not seem to be working either.
OpenAI models do not work for me in Opencode either.
Switching to Opencode until it's fixed.
After Claude Code's instability issues in the past week it seems now it's Codex CLI's turn.
CLI hangs indefinitely after prompt.
github.com/openai/codex...
Wondering what LLMs you can actually run on your hardware? 🤔
👾 llmfit — Find the best models for your RAM, CPU, and GPU
💯 Detects your system and ranks models by fit, speed & context
🦀 Written in Rust & built with @ratatui.rs
⭐ GitHub: github.com/AlexsJones/l...
#rustlang #ratatui #tui #llm
Today I caught myself writing:
“I created this project before coding agents, painstakingly hand-typing all the code.”
“Painstakingly”... that feels revealing.
On reflection, these days I mostly touch code to give examples, refactor generated code, or do the parts that are faster to write myself.
Does anyone have a favourite book on the history of nuclear weapons? I'm particularly interested in how the physicists grappled with the ethics of what they were unleashing on the world. I fear computer scientists might be the next to have such a reckoning.
Dario Amodei (CEO of Anthropic) also gave this book to his employees because he wanted them to appreciate the moral and national-security implications of building powerful technologies, comparing AI's future importance to that of nuclear weapons.
(Source one of the last Hard Fork podcasts)
I'll tell you one thing I won't be missing:
debugging GitHub Actions
You Bought Zuck’s Ray-Bans. Now Someone in Nairobi Is Watching You Poop.
Short write up of my vibe-coded site update and using Beads.
For the restyle I mainly relied Impeccable.
hanlho.com/p/experience...
Just for fun, or in case you need inspiration naming your agents, here is the list of names Codex pulls from when creating an agent:
github.com/openai/codex...
If anyone is wondering about Snipd.
It’s a European podcast service offering easy highlights which I sync to Readwise.
At first the AI features made me have some reservations but they are handy getting the gist of episodes and deciding whether to listen or not.
Moved all my tech pods to there
Latest Hard Fork episode on the recent and still ongoing developments in the US around Anthropic and OpenAI.
AI is already ethically challenging as it is, but it keeps piling up.
www.nytimes.com/2026/02/28/p...
share.snipd.com/episode/c3f7...
If code is written more and more by LLMs, the focus seems to be shifting to creating guardrails so agents can validate their own work.
Heard in: The Pragmatic Engineer - Mitchell Hashimoto’s new way of writing code
share.snipd.com/episode/d792...
Today I heard the term “harness engineering” for the first time:
Harness engineering is the practice of building tooling, tests, and automation that let coding agents execute tasks safely and reliably.
On MacOS, access all your Opencode sessions, or start new ones, using Opencode web:
`OPENCODE_SERVER_PASSWORD='your_password' opencode web --mdns --port 4096`
From your phone or another machine on your home network, access using:
opencode.local:4096
username = opencode
password = your_password
Lots of highlights for me in this article from @suksr.bsky.social
susannekaiser.net/building-fou...
Screenshot: The EU (still) wants to scan your private messages and photos. 4 member states opposing, 23 member states supporting, 0 member states undecided.
Chat Control 1.0 lets Big Tech scan ALL your messages. 👀
And the EU wants this to continue. Decision is this WEDNESDAY.
⚡ Call your MEPs now! ⚡
➡️ fightchatcontrol.eu
I’ve been writing a lot of “skills” lately.
Today I may have pushed the granularity limits a bit too far: I created one just to keep my blog tags consistent.
Probably overkill, but GPT gave me a name for it that I like: a micro-skill.
Next step: make it hexagonal.
If you haven't yet changed how you teach and evaluate thanks to AI, you probably need to. This is not something that educators asked to happen, but it is worth noting what is out there - in this case an OpenClaw tool designed to cheat.
whenever people talk about "shipping speed" they focus on 0 to 1
but what actually matters is how fast the team is still shipping a year into the project
this is why everyone seems like they "ship fast" but nothing actually seems to be getting done
Cognitive debt is a useful definition but I do not think it is new, it is now being magnified by coding agents and likely becoming more of a real risk to software products.
“The code might have been messy, but the bigger issue was that the theory of the system, their shared understanding, had fragmented or disappeared entirely. They had accumulated cognitive debt faster than technical debt, and it paralyzed them.”
margaretstorey.com/blog/2026/02...
