Detailed Introduction
Hashbrown is an open-source framework for running AI agents in the browser. It aims to bring multi-step task orchestration and external tool invocation into the frontend environment. With integrations for Angular and React, developers can coordinate model calls, page interactions, and external services on the client side in a controlled and secure manner, enabling agents that directly manipulate browser UIs.
Main Features
- Browser runtime: execute agent logic in the frontend to reduce backend dependence.
- Frontend integration: adapters for Angular and React to simplify cooperation with components and routing.
- Tools & adapters: built-in or extensible connectors for models, retrieval, and external APIs to support common LLM workflows.
- Security & control: focuses on permission boundaries and execution isolation, easing debugging and observability in the browser.
Use Cases
Suitable for scenarios that require coordinating interactive logic and model calls on the client, such as enhanced web assistants, automated form and data scraping agents, UI-focused task orchestrators, or browser-side triggers for RAG workflows.
Technical Features
Implemented in TypeScript, Hashbrown emphasizes compatibility with modern frontend build chains and component systems. Its architecture uses modular adapters and workflow orchestration as core concepts to allow inserting external models, caches, and retrieval components into browser-side execution paths. The project is open source and community-friendly.