1. Home
  2. Workshops
  3. Test-Driven Development with Coding Agents

Test-Driven Development with Coding Agents

Coding agents are fast, but without a clear contract they produce plausible code that nobody can trust. In this hands-on workshop your team learns to put tests first: you write or review the failing test, the agent makes it pass, and a senior engineer stays accountable for the result. We work on your own repository or a sandbox project.

Request without obligation

Deutsche Version

Duration
1 day
Level
Intermediate
Format
Remote or on-site, English or German
Group size
Up to 12 participants per trainer

Who it is for

  • Developers who already use a coding agent and want more reliable results
  • Tech leads who want a repeatable agent workflow
  • QA engineers who write tests alongside developers

What your team will be able to do

  • Run a red-green-refactor loop with a coding agent
  • Write tests that describe behaviour precisely enough for an agent to implement
  • Spot tests an agent has weakened or bypassed to make them pass
  • Use failing tests as stop conditions for agent runs
  • Define a team rule for what an agent may and may not change in tests

Agenda

  1. Why tests are the contractHow agents fail without a verifiable target, and why test-first gives you control over speed.
  2. Red: writing the failing testExpress requirements as small tests, either by hand or drafted by the agent and reviewed by you.
  3. Green: let the agent implementGuide the agent with focused prompts and context, and keep each iteration small and reviewable.
  4. Refactor under a safety netUse the passing suite to let the agent clean up code while behaviour stays fixed.
  5. Pitfalls and test hygieneRecognise deleted assertions, over-mocking and tests that only mirror the implementation.
  6. Team workflow and review rulesAgree on commit structure, review checkpoints and Definition of Done for agent-written code.

Prerequisites

  • Daily experience with at least one programming language and its test framework
  • Access to a coding agent (we can advise on setup beforehand)
  • A repository or sandbox project, ideally with a runnable test suite

Why apprat.io workshops

  • Taught by practising AI Code Engineers
  • Your repositories or a sandbox project
  • AI speed, human responsibility: review, security and governance included
  • ISO 9001 & ISO/IEC 27001 certified processes
  • Follow-up office hours after 30 days

Request without obligation

Tell us about your team and goals. We reply within one business day with a tailored agenda and a quote – free and without obligation.

24/7 call support in English & German:
+49 761 519 751 50

Frequently asked questions

Which test frameworks do you cover?

We work with the framework your team already uses. The principles are language-independent and we tailor examples to your stack.

Do we need an existing test suite?

No. A small sandbox project is enough, and a legacy codebase with few tests makes a good exercise for characterisation tests.

Can the agent write the tests itself?

Yes, as a draft. The workshop shows how to review such tests so that a human still owns the specification.

Fields marked * are required.

Related workshops

All workshops