Dead Code 💀

@TheShiftJournal Work book

Here’s how to actually do the things you say you’ll do

use your browser in cursor. been trying this for a bit and it feels quite magical. let us know how we can make it better

Introducing Claude Sonnet 4.5—the best coding model in the world. It's the strongest model for building complex agents. It's the best model at using computers. And it shows substantial gains on tests of reasoning and math.

The biggest application of MCP servers: building them

@leerob More flexibility on the parameters around the cursor-agent CLI: x.com/imdeadcode/sta…

What do you want to see us ship/fix in Cursor?

This is a game changer for web dev

Is there currently a way to pass permissions to a new @cursor_ai CLIR session? Similar to --settings from Claude. I know about cli.json, but I'm trying to create a collection of agents that have their own unique set of capabilities. cc @rohanvarma @leerob

Your Agent should use a File System This is a hill I will die on. Every agent can use a file system. The file system is an elegant way of representing state that your agent could read into context & allowing it to verify its work. 🧵on why and examples

@gkcs_ Both are bad advice: x.com/imdeadcode/sta…

This is bad advice. 6 books is roughly 1800 pages. That'll take forever to finish. Pick up the InterviewReady course instead, and finish it in a month. Don't give yourself impossible challenges.

@neatprompts I wish it was that simple, you either need the basics of CS and coding or a couple of years of experience as SWE as a base line at least. Please don’t pass this as easy to other folks. This “everything” from beginner to pro is far from true.

Wrote up the complete framework with code examples: deadcode.blog/blog/how-to-ev…

The game-changer? Using Claude to analyze evaluation transcripts and suggest tool improvements. Agent-optimized tools actually outperformed human-written ones.

Just read @AnthropicAI post on writing effective tools for LLM agents and it completely changed how I think about testing AI tools🧵:

@EXM7777 Image

i'm actually insane for sharing this but here we go... i built the complete AI Image Generation System: - 6-module course covering Gemini, Midjourney, GPT and Seedream 4.0 - blueprints to craft perfect prompts and start making money - 20 JSON templates for Nano Banana - 15 Midjourney SREF styles - access to Composition AI - a model trained on elite techniques to generates perfect prompts (and images) for you i'm taking this down in 24 hours because honestly this should be paid content reply "IMAGE" + retweet for access (must be following so i can send the dm)

The Claude Code SDK now supports custom tools and hooks directly in code. Additionally, we’ve refreshed all our docs with complete references and 10 new guides on how to utilize the SDK.

The frameworks will catch up eventually, but right now, custom building blocks win. Focus on solid software engineering principles with strategic AI integration. What's your experience building with AI? Are you using frameworks or custom approaches?

Take your big problem, break it down into smaller problems, then solve each using these 7 building blocks chained together. That's how you build effective AI applications that actually work in production.

After building dozens of AI apps and talking to 100+ developers, here's the uncomfortable truth: Most "AI agent frameworks" aren't being used in production. Here's what actually works for building real AI applications 🧵👇