HN
Today

A Claude Code and Codex Skill for Deliberate Skill Development

This GitHub project introduces a plugin for AI coding assistants like Claude and Codex, designed to integrate science-backed learning exercises directly into the development workflow. It tackles the often-overlooked challenge of skill degradation when using AI, encouraging developers to build expertise rather than passively accepting generated code. By offering structured reflection and practice, it aims to turn AI-assisted coding into a deliberate learning opportunity.

9
Score
1
Comments
#12
Highest Rank
6h
on Front Page
First Seen
May 14, 6:00 AM
Last Seen
May 14, 11:00 AM
Rank Over Time
261526291812

The Lowdown

The learning-opportunities skill is a plugin for AI coding assistants such as Claude Code and Codex, designed to counteract the potential for decreased learning and skill development when developers rely heavily on AI. Instead of just generating code, this tool integrates evidence-based learning exercises directly into the development process, aiming to foster true expertise.

  • Core Functionality: The skill proposes optional 10-15 minute learning exercises during significant architectural work (e.g., new files, schema changes, refactors). These exercises leverage techniques like prediction, generation, retrieval practice, and spaced repetition.
  • Counteracting AI Risks: It addresses concerns such as the 'generation effect' (reduced active processing), 'fluency illusion' (overestimating understanding of generated code), lack of 'spacing effect' (due to continuous coding), and diminished 'metacognition' and 'retrieval practice' when AI provides complete answers.
  • Interactive Exercises: Claude pauses and waits for user input during exercises, intentionally pushing against its default to provide full answers, thereby encouraging active mental effort. Examples include 'Prediction → Observation → Reflection', 'Generation → Comparison', 'Trace the path', 'Debug this', 'Teach it back', and 'Retrieval check-ins'.
  • Supplementary Tools: The repository also includes learning-opportunities-auto for optional post-commit prompting and orient to generate repository orientation lessons, based on research into expert developer navigation.
  • Scientific Foundation: The entire approach is grounded in well-established findings from learning science and empirical research on developer thriving and AI-assisted workflows.
  • Team Adoption & Customization: A MEASURE-THIS.md playbook is provided for teams to track the experiment's visibility and impact, offering validated survey items and guidance. The skill is also highly customizable, allowing adjustments to trigger conditions, exercise content, and integration with existing knowledge.

Ultimately, learning-opportunities aims to transform the AI coding experience from passive consumption into an active, deliberate process of skill acquisition, ensuring developers continue to grow their capabilities alongside their AI counterparts.