Overview

• Anthropic has enabled a “Learning” preset in Claude.ai’s style menu to guide users through Socratic questioning rather than providing direct answers.
• Claude Code now offers two teaching modes—Explanatory, which narrates coding decisions, and Learning, which pauses with “#TODO” prompts for developers to fill in code.
• Developers can activate these modes by updating Claude Code, using the /output-styles command and selecting from Default, Explanatory or Learning, or create custom styles via the Output Styles API.
• The learning behaviors are powered by editable system prompts instead of fine-tuned models, allowing fast iteration but leading to occasional inconsistencies across conversations.
• Existing campus partnerships with Northeastern University, the London School of Economics and Champlain College bolster Anthropic’s push against OpenAI’s Study Mode and Google’s Guided Learning in the education market.