Most Claude Code sessions go wrong in the first message.
Not because the model is weak, but because the starting prompt is vague: “fix this bug” or “add tests.” With nothing else to go on, the model starts editing before it has looked at the repository, and you spend the next ten minutes steering it back.
What changes the outcome is the order of the work. Inspect the repo first. Learn the existing conventions. Plan before editing. Make focused changes, add or update tests, run real commands to verify, then summarize what changed. A prompt that builds that sequence in means you don’t have to remember to ask for it every time.
I put together a set of 2,000 prompts around that workflow, covering areas like debugging, testing, security, architecture, APIs, and cloud. They come as JSON, CSV, and JSONL, so you can browse them, filter them, or feed them into your own scripts.
Plan before editing is the step I see skipped most. When I build with Claude, asking it to restate the repo's conventions back to me before any change catches most wrong assumptions early.
Restating the conventions back is a smart check, because it makes the model show its assumptions before it touches any code. I’d add one more step: have it name which files it plans to change and why. If that list looks wrong, the plan is wrong, and it’s much cheaper to catch that before the edits.
The colorful spines on the wood shelf under the big TOKYO STYLE title feel like a real bookstore. When a book slides out, does the shelf close the gap or leave a space?
I think we're heading towards a stage where everyone will have their own unique tools and workflows, which they won't even bother sharing or selling, knowing others can simply reverse-engineer and adapt them. Feels like we're entering an era of truly unique creativity.
I built a portfolio that turns its pages like a real notebook. Scroll to flip, click to open a project. Leather cover, typewriter ink, little stickers, and easy to make your own. 📓
link