Greg Leuch

🎯 5 Practical Claude Prompts for Writing and Editing Set up a Claude Project for each story, use web search for real-time intel, and Artifacts for collaborative editing - Claude supports your entire editorial workflow. Here are 5 prompts to use daily: Research: "I'm writing about [story/topic]. Give me the essential context: key players, timeline of events, different stakeholder positions, and 3 angles that others might miss." Sources and quotes: "Help me find expert sources for [topic]. Suggest 5 credible voices with different perspectives, their likely viewpoints, and smart questions to ask each. Use Claude's web search to find the most recent developments and cross-reference multiple sources." Make complex topics accessible: "Take this complex [policy/study/situation] and help me explain it clearly. Create analogies, identify the human impact, and suggest compelling ledes that make readers care about why this matters." Adapting content for different audiences: "I have this core story: [your content]. Help me adapt it for different formats: breaking news alert, in-depth feature, social media thread, and newsletter summary. Optimize each for its specific audience and platform." Tight deadlines: "I have [time limit] to write [type of piece] about [topic]. Create this as a Claude Artifact so I can easily edit and iterate on the draft. Give me a sharp outline, key points to investigate, and 3 compelling opening paragraphs to choose from. Include questions to verify key facts." What's your go-to prompt? Drop it below 👇

76

6 Comments

Here’s a little sneak peek at our Firefox icon system design guidelines. These visual rules cover style, base grid, key shapes, metrics, and sizing, along with recommended practices and things to avoid (do’s and don’ts) 📐✨

266

7 Comments

Couldn't resist testing Gemini Studio with Gemini 3 for a new generative UI demo. This time, everything is generated - both the images and the text within them. The setup: Two views working in tandem. A blueprint editor where you define content blocks (headers, content sections, buttons). This blueprint feeds directly into the image generation. Then the UI view shows the rendered outcome, with interactive elements like buttons overlayed on the generated images. What impressed me: Gemini 3 handled every iteration flawlessly. No wrestling with prompts, no failed generations. It just worked. This opens up some fascinating possibilities. Think infinite zoom interfaces with layered images. Or seamless blending of generated images and video within the same UI. As image generation continues improving, we're moving past "AI-generated assets" into "AI-generated interfaces." The constraints that have defined web design for decades - grid systems, component libraries, design tokens - might become optional rather than mandatory. You could describe an interface and have it materialize, pixel-perfect, with all interactive elements in place. Looking forward to pushing this further. #GenerativeUI #AIDesign #InterfaceDesign

35

7 Comments