Abstract
A cloud-enabled, chat-controlled architecture provides remote AI agent access to a local desktop design application through a protocol bridge. A plugin running inside the desktop application connects to a locally executed bridge service using WebSocket and executes commands via an internal plugin API with deep read/write support, including complex assets such as images and fonts. A cloud-hosted agent receives natural-language instructions from a chat platform, retrieves context from an external knowledge base and optionally a brand asset pipeline, generates structured tool commands subject to design-system constraints, and sends the commands as HTTP requests to the local bridge. The bridge translates HTTP requests into WebSocket messages for the plugin and correlates plugin responses back to HTTP responses, enabling bidirectional workflows including user-triggered actions from the plugin UI.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Anonymous, "Protocol Bridge Architecture for Chat-Directed Cloud AI Agent Access to Local Application Interfaces with Knowledge Base Integration", Technical Disclosure Commons, ()
https://www.tdcommons.org/dpubs_series/10646