Slack engineering has launched a new testing methodology known as agentic testing, which leverages artificial intelligence (AI) agents to enhance end-to-end testing resilience in dynamic software environments. Traditional end-to-end tests often fail due to changes in user interfaces or services, rather than actual functional regressions, leading to increased maintenance burdens. Agentic testing shifts responsibility from static scripts to AI-driven agents that execute workflows based on higher-level intents rather than fixed sequences of actions. Tests are articulated as objectives, allowing AI agents to interpret intent and interact with applications through UI or API surfaces. The agents dynamically evaluate the application state at each step, adapting their actions in response to minor changes. The workflow begins with a test intent passed to the agent layer, which then plans, executes actions, observes results, and iterates until the objective is achieved or a stopping condition is met. Currently, Slack engineers note that agent-driven testing is more suited for targeted debugging and exploratory testing rather than frequent execution in continuous integration pipelines due to cost considerations. The system incorporates constraints to govern agent behavior, including limits on allowed actions and conditions for execution termination.