Geoffrey Litt on X: "real prompt i used today: ``` Can you write out another Markdown document called `Implementation Guide`? This is similar to the document you wrote, but it's more of a guide for a junior developer to reimplement the PR from scratch. I want you to give an abbreviated version of" / X
To view keyboard shortcuts, press question mark View keyboard shortcuts Home Explore 3 Notifications Chat SuperGrok Premium Lists Bookmarks Communities Profile More Post floguo @floguo Post See new posts Conversation Geoffrey Litt @geoffreylitt · Nov 21 I cannot emphasize enough how much I prefer this "tutorial doc + build-it-yourself" coding workflow to the typical "ugh" feeling of reviewing huge agent PRs. You can try it right now and see for yourself: 1) Instead of having Claude Code make a PR, ask it to output a Markdown Show more Quote Geoffrey Litt @geoffreylitt · Nov 21 Replying to @geoffreylitt on to my next task! following the tutorial guide claude precooked 25 96 1.5K 179K Geoffrey Litt @geoffreylitt · Nov 21 addenda: - point Cursor at the implementation guide as context, I'll often make a new composer thread and say something like "plz build step 1.3 in http:// guide.md and then i'll validate it" - claude code is much better than codex cli at writing super-in-depth Show
Geoffrey Litt @geoffreylitt Replying to @geoffreylitt real prompt i used today: ``` Can you write out another Markdown document called `Implementation Guide`? This is similar to the document you wrote, but it's more of a guide for a junior developer to reimplement the PR from scratch. I want you to give an abbreviated version of the background and then walk through various milestones, ideally with some verifiable behavior at each step. Each milestone has to be verifiable in some way, whether it's tests or visible behavior. Write it as a tutorial, specifying what changes to make and how to test
Explore this link on the map →related reading
- llm-wiki · GitHubgist.github.com
- Best practices for Claude Code - Claude Code Docsanthropic.com
- Impeccable: The missing upgrade to Anthropic's impeccable skillimpeccable.style
- Cookbookcookbook.openai.com
- Prompt guidance | OpenAI APIdevelopers.openai.com
- GitHub - brexhq/prompt-engineering: Tips and tricks for working with Large Language Models like OpenAI's GPT-4. · GitHubgithub.com
- Prompt engineering | OpenAI APIplatform.openai.com
- How to write better prompts for GitHub Copilot - The GitHub Bloggithub.blog
- Using CLAUDE.MD files: Customizing Claude Code for your codebase | Claudeclaude.com
- Using CLAUDE.MD files: Customizing Claude Code for your codebase | Claudeclaude.com
- Semantic Commit Messages · GitHubgist.github.com
- Figma Make Prompt Architect | Prompts1777.studio