Elixir developers know metaprogramming: writing code that writes code. What if we applied the same thinking to AI prompts?
Metaprompting is creating prompts that write and persist other prompts. Just as Elixir’s metaprogramming generates context-appropriate code, metaprompts generate context-appropriate instructions for AI assistants. This talk draws high-level comparisons between programming in Elixir and programming with Claude.
Through examples like scout/plan/build/QA workflows, you’ll see how the same patterns that make Elixir code elegant—composition, clear responsibilities, reusable abstractions—also make AI-assisted development reliable. Learn to think about prompt engineering as a programming discipline by applying principles you already know: small focused pieces, context awareness, and generated artifacts that persist across sessions. developer joy.
OBJECTIVES To teach you:
- Thinking about prompt engineering as programming
- When to use templates vs. metaprompts
- Building multi-phase workflows with generated prompts
- Persisting generated prompts as project artifacts
- When metaprompting creates leverage vs. complexity
AUDIENCE
- Elixir developers interested in systematic AI-assisted development approaches. No advanced metaprogramming knowledge required.