Vai al contenuto
AI.info

events

Oracolo del Test - Beyond Autocomplete: Architecting Black-Box Agentic Testing

Ludovico Besana sui test agentici a scatola nera: agenti LLM che controllano un’app in esecuzione tramite Playwright MCP e Chrome DevTools MCP, con una demo e una panoramica su costi, latenza e misure di sicurezza.

Oracolo del Test - Beyond Autocomplete: Architecting Black-Box Agentic Testing

In inglese

Starts
2026-09-30T17:00:00+00:00
Organizer
Oracolo del Test
Location
Sala vetrata Bassini, via Bassini 6, Milano, Italy

An evening session hosted by Oracolo del Test at the Politecnico di Milano's Sala vetrata Bassini. Ludovico Besana presents Black-Box Agentic Testing: instead of generating test code, AI agents drive a running application themselves through the Model Context Protocol, using browser and testing tools in a Think → Act → Observe loop. A live demo wires Playwright MCP and Chrome DevTools MCP to an LLM that explores a real application without scripts. The talk also covers the limits of the approach — token costs, latency, non-deterministic and flaky runs — and the guardrails needed, plus where agentic testing fits into a quality strategy.

Event website