Short Coding Course — South Africa
Context Engineering, Claude Code, GitHub Copilot & AI Code Review
AI won't replace developers — but developers who use AI well are pulling ahead of those who don't. This hands-on course shows working developers how to use GitHub Copilot and Claude Code professionally on real codebases: giving the AI the right context, testing and reviewing what it writes, automating code review in CI, and keeping company code safe.
What's happening behind the assistant: tokens, context windows, why models invent APIs that don't exist, and how model choice affects quality and cost. Choosing the right tool for the job — from inline completion to a full coding agent.
Why context matters more than clever prompts. Writing specs and to-do lists before code, project instruction files (CLAUDE.md, AGENTS.md, Copilot custom instructions), giving the AI the right files, and knowing when to start a fresh session.
Inline completions, Copilot Chat, agent mode in VS Code and JetBrains, and assigning issues to the Copilot coding agent. Custom instructions so Copilot follows your team's conventions.
Installing and configuring Claude Code, planning before coding, multi-file changes, reviewing diffs, and using Git checkpoints so you can always roll back.
Custom slash commands for repeatable tasks, agent skills, hooks that run your tests and linters automatically, subagents, and connecting MCP servers to your database, GitHub and documentation.
Debugging from stack traces, refactoring legacy code, getting up to speed on an unfamiliar codebase, designing database schemas, and generating documentation that stays accurate.
Generating unit and integration tests, test-first prompting, catching hallucinated functions and edge cases, and making sure tests check behaviour rather than just pass.
Reviewing AI output critically, prompting for security and performance review, PR summaries, and adding an automated AI review step to your CI pipeline with GitHub Actions.
Prompt injection through files, web pages and MCP servers; keeping secrets out of prompts; agent permissions and sandboxing; licensing and IP; POPIA and what may be sent to cloud AI services; local models as an alternative; and writing a team AI usage policy.
Calling the OpenAI and Claude APIs from your own scripts with structured output to build a small internal tool. Capstone: apply the full workflow to a real codebase — a feature with tests, AI review, CI and deployment — and demo it.
This course is for working developers. Make sure you're ready before enrolling.
At least 2 years of writing code professionally, in any language. You need to be able to judge whether AI-generated code is good.
Comfortable with branches, commits, pull requests and running commands in a terminal — AI coding agents work through both.
Bring a project you know well (or use our provided codebase). Applying AI to real, existing code is the heart of the course.
Tool access: A GitHub account and access to GitHub Copilot and Claude Code during the course. We confirm licence options with you when you book.
Some AI tool experience: Prior use of ChatGPT, Copilot or similar is helpful but not required.
Optional: build apps fast with Cursor, Windsurf and Lovable, and learn the core AI-coding habits.
Ready to take this course! Use Claude Code, GitHub Copilot and AI code review professionally on real codebases.
Next: put AI inside your own products with LLM APIs, RAG, agents and MCP servers.
You already use Copilot or ChatGPT, but the results are hit-and-miss. Learn the workflow that makes AI output reliable on real code.
Agree on shared instruction files, review standards and an AI usage policy so the whole team gets consistent, safe results.
Set up Claude Code, Copilot and AI code review in CI for your team — and know where the risks are.
Take the habits you learned building new apps and apply them professionally to large, existing codebases.
All courses are taught by Code College’s senior development trainers — practitioners with real-world industry experience, backed by over 20 years of professional developer training since 2004.
You write code from day one. Every concept is reinforced with real exercises and a practical project you keep.
Join 100% live online from anywhere, or hybrid — one or two days a week at our Midrand campus and the rest live online. Full-time face-to-face attendance can be requested, and is possible when an onsite group class is running.
Receive a Code College certificate on completion — recognised by employers across South Africa.
No. The course works with any language. You need at least 2 years of professional development experience and comfort with Git and the command line. The short API module uses simple scripts that you can follow in Python or JavaScript.
GitHub Copilot (including agent mode and the Copilot coding agent), Claude Code with commands, skills, hooks and MCP servers, and the OpenAI and Claude APIs for automating your own workflow. The capstone applies the full workflow to a real codebase.
It can be — if you set it up properly. A full module covers prompt injection, keeping secrets out of prompts, agent permissions, licensing and IP, POPIA and what may be sent to cloud AI services, local models as an alternative, and writing a team AI usage policy.
Vibe Coding is about building new apps fast with Cursor, Windsurf and Lovable. AI Tools for Developers is about using AI professionally on existing codebases — context engineering, Claude Code, GitHub Copilot, testing, AI code review in CI and security. For building AI features into your own products, see Building with LLMs.
Yes. Groups can book a 20-hour condensed version (2.5 days) covering the core modules: Context Engineering & Prompting for Code, GitHub Copilot in Practice, Agentic Coding with Claude Code, Extending Claude Code, and AI Code Review & CI Integration. Contact us for a group quote.
Individuals join live online from anywhere, or hybrid — one or two days a week at our Midrand campus and the rest live online. Group courses run live online or hybrid, with fully onsite delivery at our Midrand campus optional, on request. Teams of 2–4 save 10% and teams of 5 or more save 20% per person (same class dates, paid in full upfront). Contact us for a group quote.
Yes. Code College is a B-BBEE Level 4 contributor and training spend qualifies as skills development under the B-BBEE scorecard. The course fee also counts toward your SDL (Skills Development Levy) spend.
Book the AI Tools for Developers course today — full-time (5 days) or part-time. Johannesburg and online.