The WebRider framework aims to provide persona-conditioned intent controllers for Live-Web assistance by formalizing delegated policies as intent contracts. This contract includes goals, constraints, evidence obligations, answer form, and task-local persona controls. WebRider employs a hierarchical architecture: a top-layer controller maintains the contract, a middle layer realizes intentions as guarded executable actions, and a tool layer executes these actions via browser, search, and maps tools. The RiderBench benchmark evaluates this design on 4,096 live-web contracts, auditing both the internal contract state and the visible user experience to determine if a rollout preserved its policy and if the steps were persona-consistent. The guarded middle interface also serves as a high-quality training signal; an 8B action-policy model trained through this interface outperforms executable-only baselines under a fixed controller. By making the browsing path a first-class object, WebRider enables a system that is auditable, human-judgeable, and learnable without conflating action realization with final-answer decisions. Blogger's Review: The WebRider framework significantly improves the reliability and persona consistency of Live-Web assistants by explicitly defining the intent and policy of delegated tasks, providing a foundation for auditable, human-evaluable, and learnable Live-Web assistant systems.