AI-Assisted Coding | Session 1: Code Generation — Safe Prompts and Verified Change
AI-Assisted Coding | Session 1: Code Generation — Safe Prompts and Verified Change
Share this event
AI-Assisted Coding | Session 1: Code Generation — Safe Prompts and Verified Change
We offer a 50% discount to students; please complete this form to avail the discount.
How do you turn a vague coding request into an AI-generated change you can prove is correct?
This session introduces safe AI-assisted code generation, covering how large language models generate code and where they fail. You will learn to construct prompts using clearly defined tasks, constraints, examples and acceptance criteria.
The session also covers data classification, secrets, and intellectual property, along with verification through linting, pytest, security checks, and human review.
You will explore requirements, definitions of done, task decomposition, repository context, test-driven debugging, safe refactoring, Git diff review and pull-request summaries.
This session is the first of a three-session professional training programme:
REGISTRATION FOR ALL 3 SESSIONS IS 150$ ONLY.
What You Will Learn
How LLM code generation works and where it fails
Prompt construction using tasks, constraints, examples and acceptance criteria
Data classification and protection of secrets and intellectual property
Verification through linting, pytest, security checks and human review
Requirements and the definition of done
Task decomposition and repository context
Test-driven debugging and safe refactoring
Git diff review and pull-request summaries
Why This Matters
AI-generated code should not be accepted simply because it appears plausible. Engineering teams need bounded tasks, controlled context and explicit acceptance criteria, followed by verification and human review.
This session establishes the verify-before-accept discipline that will be applied throughout the programme.
Ideal For
Software engineers
Technical leads
QA engineers
Software architects
Final-year and postgraduate learners with programming and Git experience
Training Format
4-hour live online training session
Concise teaching blocks
- Four live demonstrations:
Demo 1: Turn a weak coding request into a constrained prompt with examples and acceptance tests.
Demo 2: Remove unsafe context and verify generated code with linting, pytest and human review.
Demo 3: Convert a feature request into a bounded task plan with only the relevant repository context.
Demo 4: Reproduce a failing test, guide the repair and review the Git diff.
A ten-minute break and five-minute quiz exactly at the two-hour mark
Practical exercises using a real sample repository
Tools and Technologies
Claude Code or Codex, Python, Git and pytest.
This session is Session 1 of the three-session AI-Assisted Coding and Code Generation Programme.