Building an Intern
A developer describes building an AI-powered software engineering intern that autonomously breaks down tasks, writes code, runs it, and fixes errors using large language models.
Background
- **Cra.mr** is the blog of Craig Mod, a writer and designer who often explores the intersection of craft, technology, and slow/thoughtful processes.
- This piece is about creating a custom AI coding assistant (an "AI intern") — not a generic chatbot, but a deeply personalized tool trained on the author's own work patterns, codebase, and preferences.
- The core idea: instead of using off-the-shelf AI coding tools (like GitHub Copilot or ChatGPT), you build your own by curating specific documents, rules, and examples that reflect how *you* actually work. The "intern" learns your style and handles repetitive or scutwork tasks.
- This reflects a growing movement among developers who find generic AI assistants too noisy or misaligned, and instead invest time in "prompt engineering" and custom toolchains to get more reliable, taste-aligned help.
- The broader significance: it shifts the conversation from "AI replaces workers" to "AI as a highly personalized apprentice that amplifies your existing judgment and craft."