Sabitlenmiş Tweet
My first iOS app is live!
I got tired of losing track of which prompts actually worked across different models and platforms. So I built PromptFolio. apps.apple.com/us/app/promptf…
English
michael
1.3K posts
michael
@mikelaaron
Update: Couldn't stop losing my best AI prompts. Talked AI into helping me build PromptFolio. Still can't draw...
Arlington, VA Katılım Nisan 2019
2K Takip Edilen144 Takipçiler
@capitalweather We so badly need the rain in Arlington VA. But we always fall victim to the DC hole that pushes storms north and south 😭
English
So far, most storminess south/east of Beltway. Heavy storm around Upper Marlboro meandering.
Additional scattered storms likely develop or move in from west. Example of what radar may look like late eve here.
Updated forecast at capitalweather.com or Capital Weather app.
English
@Shpigford Surprised it wasn't triggered by you making a game related to organisms. Mine was 😭
English
super confused on fable. been using it for the past few hours to build souplings.fun and it hasn't triggered any security downgrades and also is barely using any tokens.
which...great. but also 🤔
English
@JJEnglert Sonnet 5 keeps telling me to just paste my API keys in the chat- whereas previous versions of the model, and other company's models, all will go out of their way to find a save way to add API keys to a project.
English
Sonnet 5 vs. Opus 4.8
The price gap is real:
Sonnet 5 runs $2/$10 per million tokens through August, then $3/$15.
Opus 4.8 is $5/$25. That's roughly 40-60% cheaper depending on when you're reading this.
Use Sonnet 5 when:
- Daily, busy work: slides, internal docs, first drafts, status write-ups
- Automations or agents running at volume, where the price gap compounds fast over hundreds or thousands of runs
- You want speed over deliberation. It gets to the point instead of thinking for 5 minutes first
- Basically anything outside pure engineering, especially if your team actually feels the API bill
Use Opus when:
- The task needs restraint: refactors, sensitive client work, anything where a model that's too confident can do real damage
- It's genuinely high-stakes, where a mistake costs more than what you'd save on tokens
- You need the deepest, most deliberate reasoning available and you're fine paying for it
- You haven't tested Sonnet 5 on that exact task type yet and the downside of being wrong is high
Still testing. If you've run your own Opus comparison, send it my way.
Claude@claudeai
Introducing Claude Sonnet 5, our most agentic Sonnet yet. It makes plans, uses tools like browsers and terminals, and runs autonomously at a level that just a few months ago required larger and more expensive models.
English
It would be very useful to understand more about the government safety concerns associated with frontier AI releases so we could (a) know what risks everyone will face if/when open source reaches Mythos class & (b) whether they are doing enough or too much to prevent those risks.
English
What if Claude Code could improve your prompt before it's sent?
A lightweight "Refine" concept that uses the project context Claude already has to turn rough ideas into clearer tasks.
The best clarification is the one you never need.
English
I have this tool Claude and I made: just type "twit" in your terminal followed by the URL of the X post and it saves it to your location of choice (mine is Obsidian): gist.github.com/mikelaaron/a33…
Ole Lehmann@itsolelehmann
luckily bookmark rot is an easy problem to fix now here's how to turn every X bookmark you've ever saved into a second brain your agent has full context on: 1. export your bookmarks. i use twitter-web-exporter (free userscript) or the BookmarkSave extension. you get one file with every bookmark + the full text + the author + the link 2. drop that file into a folder. if you already run an llm wiki / obsidian vault, drop it straight in so your bookmarks join the rest of your knowledge 3. point your agent at the folder (claude code, codex, hermes, whatever you run) and tell it: "read this export and turn every bookmark into its own markdown note with the original link and a couple of topic tags" that's it, your agent has read all of it. now you can ask "what have i saved about pricing" or "pull everything i bookmarked on claude code" and it answers across the whole pile takes maybe 10 minutes after that they actually get used, and every new bookmark folds into the same brain instead of rotting in a tab you never open again
English
In case anyone on a mac needs a quick way to make a new markdown file in folder using right-click, I created: github.com/mikelaaron/mdM…
English
Update on Agent Time Machine: the snapshot and restore engine works. Agents checkpoint, agents restore, the MCP tools are solid across Claude Code, Codex, and Gemini.
michael@mikelaaron
Before I go deep, I want to know: • Does this problem resonate? • Would your agent benefit from self-awareness tools? • What am I missing? Roast it, improve it, or tell me it’s dumb. #buildinpublic
English
@LyalinDotCom and @OfficialLoganK Sorry to bother you- hope things are well. I don't know where to turn for help with Antigravity. I'm an Ultra subscriber but every single prompt fails. Anyone you can offer to help?
Ok- love you. bye.
English
@jocarrasqueira @antigravity @_anshulr @geekyouup I repeatedly can’t get it to work- even as an ultra subscriber- it gives me an error and just says to retry the prompt. just wish I could get help
English
The @antigravity team keeps rolling out new features and your feedback is key!
What can the team improve/ do next?
Cc: @_anshulr @geekyouup
Google Cloud Tech@GoogleCloudTech
With Antigravity, you collaborate with your agent using Google Docs-style commenting. Highlight specific sections of your app or data tables to provide contextual feedback and refine features in real time. Learn how to collaborate with Artifacts, here → goo.gle/4kW3y4z
English
@mikelaaron git stash before every ai session became muscle memory for exactly this reason
the worst is when it 'helpfully' standardizes your error handling across 30 files... technically correct but now you have a 2000 line diff to review for a one-line bug fix
English
Before I go deep, I want to know:
• Does this problem resonate?
• Would your agent benefit from self-awareness tools?
• What am I missing?
Roast it, improve it, or tell me it’s dumb.
#buildinpublic
English