Breaking large tasks into steps with Goblin Tools
Bram De Buyser built Magic ToDo to turn a large task into smaller actions you can edit and check off.
See what people actually built with AI: the outcome, the stack, the workflow and the evidence. Find an idea you can adapt.
Bram De Buyser built Magic ToDo to turn a large task into smaller actions you can edit and check off.
Be My Eyes made image descriptions and follow-up questions available inside its app, with input from blind and low-vision testers.
Techno Tim combined Paperless-ngx with local models for metadata and vision OCR, then showed the results on real scans.
Adam Shoqar at Learn It Live connected a website chatbot to a Zapier Table, leaving people to handle more specific support problems.
Angela Ferrante and the Laudable team connected call transcripts, ChatGPT and automation so follow-ups could reflect each customer's goals.
Sasha Kasiuha's team used generated images and Runway Gen-2 to make visuals shown during Madonna's Celebration Tour.
Simon Willison used Claude 3.7 Sonnet and a small script to publish short, source-linked descriptions for 78 personal web tools.
Daniel Miessler published Fabric-generated notes from a long Dwarkesh Patel interview, with ideas, quotations and references organized into sections.
Oxford PharmaGenesis researchers used Elicit for structured, source-linked extraction while keeping human validation in the workflow.
Jay at Hound Hill Homestead connected camera snapshots, Gemini and Home Assistant to a dashboard and phone notifications.
A focused editorial collection, not a popularity ranking. Each entry links to the original builder’s work and identifies what its evidence supports. Research checked 2026-10-29.