Course · 7 chapters

Tests de code pour le développement agentique

Stop Claude Code and Codex from shipping bugs by wiring static analysis, coverage, mutation, pre-commit hooks, and property-based tests. 7 chapters, practitioner, for engineers.

Paidpractitioner7 chapters142 minEnglish + 6 languagesCertificate on completion

What you'll be able to do

  • Design test strategies that catch real bugs
  • Wire static analysis into agent loops
  • Build pre-commit guardrails for AI agents
  • Run mutation testing on your suites
  • Review AI-generated tests with confidence
  • Apply property-based and fuzz testing

What's inside

  1. 1
    Tests de code pour le dev agentique : commence ici

    Une orientation de 12 minutes sur le parcours Code Testing — les couches de feedback écriture-vérification-correction qui empêchent les agents de codage IA de tricher, et l'ordre dans lequel les apprendre.

    12 min
  2. 2
    Les fondamentaux du testing

    Maîtrise la stratégie de test, le testing trophy, et comment écrire des tests automatisés qui attrapent vraiment les bugs.

    18 min
  3. 3
    Analyse statique pour le développement agentique

    Connecte Ruff, Biome, Pyrefly, ty, Semgrep et CodeQL comme sources de feedback sub-seconde pour Claude Code et Codex.

    20 min
  4. 4
    Garde-fous Pre-Commit pour les agents IA

    Connecte le formatage, le lint, le type-check et les tests dans un pipeline de hooks en couches qui force Claude Code et Codex à corriger avant de commiter.

    20 min
  5. 5
    Couverture & Tests de Mutation

    La boucle de feedback à deux outils qui te dit où les tests manquent et si ceux que tu as fonctionnent vraiment.

    22 min
  6. 6
    Génération de tests par l'IA

    Des prompts one-shot aux boucles pilotées par le feedback — pilote, supervise et révise les systèmes qui écrivent désormais tes tests.

    28 min
  7. 7
    Tests par propriétés et fuzzing

    Dépasse les exemples choisis à la main. Spécifie des invariants, laisse un générateur traquer les contre-exemples, et associe cette discipline à des agents qui proposent les propriétés que tu as manquées.

    22 min

Frequently asked questions

What will I learn in this code testing course?
You build five feedback layers that catch bugs in AI-written code: a test strategy with automated tests, static analysis using tools like Ruff, Semgrep, and CodeQL, layered pre-commit hooks, coverage paired with mutation testing, and AI-driven test generation plus property-based and fuzz testing. Every layer is shown applied to Claude Code and Codex output.
Who is this path for?
It is built for software engineers who use AI coding agents like Claude Code and Codex and want guardrails that keep generated code correct. It sits at a practitioner level, so it assumes you already write and run code day to day.
Do I need prior experience or to know how to code?
Yes. This is a practitioner-level path for engineers, not a beginner intro. You should be comfortable writing code, running a test suite, and working at the command line before you start.
How long is the course and is there a certificate?
The path has 7 chapters totaling about 142 minutes, roughly 2.4 hours, including a 12-minute orientation chapter. You earn a certificate of completion when you finish all chapters.
Is this course free?
No, this is a paid path included with an AI Academy subscription. It covers static analysis, coverage, mutation testing, pre-commit hooks, AI test generation, and property-based testing across all 7 chapters.

Earn a certificate

Complete all chapters to receive your certificate of completion.