AI Coding Assistants

Introduction

Coding assistants (Copilot, Cursor, Cody) combine inline completion + chat + codebase-aware retrieval.

Beginner analogy: Like pairing with someone who's read your codebase.

Understanding the topic

Core concepts to understand:

• Inline: code + comments as context.
• Chat: slash commands + multi-file workspace.
• Codebase RAG: search the repo for relevant files.
• Persistent rules: .cursorrules, copilot-instructions.md.

Syntax reference

Visual workflow / architecture:

User asks → retrieve relevant files (RAG) → prompt with context → generate → validate → apply

Real-world use

GitHub Copilot has 1M+ daily users; Cursor's revenue grew 100×+ in 2024.

Best practices

• Keep repo-level instructions short & specific.
• Pair RAG with file-graph awareness.
• Run tests automatically on suggested edits.

Hands-on exercise

Interview preparation — practice these questions:

• Q1. How do coding assistants stay codebase-aware?
• Q2. Difference Cursor vs Copilot?
• Q3. Repo-level instructions — what to include?
• Q4. Auto-test on suggestions?