latest
Your CLAUDE.md is probably wrong, and here's how Anthropic's engineers actually structure theirs
Your file is probably doing more harm than good
I stopped asking AI to build features and started asking it to remove code
It just kept saying yes, so I started saying no
Claude Code's sub-agents solve one problem and create three others
Isolation giveth, and isolation taketh away.
I put ChatGPT Work up against Claude Cowork, and the winner wasn't who I expected
Let the (agent) wars begin
Claude 'lost' hours of my chat history overnight, and I found it hiding on an invisible branch
Your missing chats (probably) aren't gone
I added one line to my Claude Code prompts, and the difference was night and day
One line is all it took
I created a second brain with Claude – and this is how I structured it
My blueprint for an AI brain.
I ditched Claude Code after giving my local LLM filesystem access
One MCP, and I don't have to let cloud-based tools handle my files anymore
I ditched Notion entirely after adding one file to my Claude Code setup
One file can replace the part of Notion you're paying for
Claude orchestrating Codex agents is the workflow I didn't know I needed for coding
Trust, but verify (and let the agents verify too)
I ditched Claude Code for a local model and a cloud supervisor, and it's twice as fast
The cloud plans, Qwen does chores.
Mistral's Vibe does what ChatGPT, Claude, and Gemini charge for, and it's free
The big three aren't the only ones in the game
I let Claude manage my PC overclocking remotely, and it actually worked
KVM-over-IP and a little LLM magic was the overclocking secret I needed
I replaced Claude Code with a team of tiny specialists, and the boring setup worked better
And the setup doesn't cost me anything.
I replaced Claude Design, Figma Make, and Replit with an open-source local-first alternative, and it saved me a fortune
Vibe design is expanding, and open-source is right on its tail
I gave Claude read access to my Obsidian vault, and it found connections I'd forgotten I made
Years of notes were hiding valuable insights in my Obsidian vault until Claude connected them in ways I never expected.
I used Perplexity Deep Research for a month, but Claude quietly crept into my workflow
Perplexity's Deep Research works ok, but I found that Claude is more powerful.
I let Claude clean up my Windows install, and it found bloat I didn't even know existed
Having a large boot disk is a curse and a blessing
I gave my local LLM an escape hatch to Fable 5, and it solved the problems I couldn't
Your local LLM can use some help.
Claude plus a local LLM cuts my AI costs in half, and I'm never going back to cloud-only
You can count on Claude for many things, except saving your money