Shimeles teaches Claude Code as the agentic coding tool he switched to from Cursor, with a full walkthrough of why Claude Code is outperforming other AI coding agents and what it's like to ship a real app with it from start to finish. He frames Claude Code as the more flexible primitive for production work, and his 'Claude Code Replaced Cursor for Me' framing is a direct, public switch endorsement.
Person profile
Michael Shimeles's AI stack.
Founder; Software engineer at Independent.
Tools Michael Shimeles has publicly discussed or demonstrated using. Curated by Magpie's editorial team from podcasts, posts, and public stacks.
About
Michael Shimeles runs the Ras Mic YouTube channel and a Toronto-based software studio and AI consultancy. He's a full-stack engineer with production apps that have processed $2M+ in transactions. Content covers Cursor, Claude Code, MCP and agentic engineering for software builders.
Channels
Tools they advocate for
- Claude Code4.0DirectCoding
- Cursor4.9DirectCoding
Shimeles teaches Cursor as the inline AI editor in his earlier content, with deep walkthroughs of frameworks and prompting strategies for getting Cursor to ship production code. His later content is more honest about pushing past Cursor toward Claude Code for agentic builds, but he treats Cursor as the right starting point for engineers new to AI coding.
- Claude4.7DirectLLM chatbot
Shimeles uses Claude alongside V0 and Cursor in his web-app build workflow, walking through how the three tools chain together: V0 for component scaffolding, Claude for nuanced reasoning, Cursor for inline edits. He treats Claude as the heavy-lift LLM for builds where output quality matters more than raw speed.
VideoJan 15, 2025
Last researched May 10, 2026
More people on Magpie AI
See all people →Editorial note
These tools have been mentioned by the people featured in podcasts, interviews, posts, and other public media. Source links and citations are included wherever possible. Magpie AI has no direct relationship with the people on these pages. Spot a missing tool or a wrong attribution? Suggest a correction →