Block has introduced Builderbot, an AI tool designed to coordinate multiple agents across their entire codebase directly within Slack, which streamlines code development significantly. Builderbot can execute tasks such as researching, planning, coding, opening pull requests, monitoring continuous integration, and incorporating feedback, enabling the merging of around 1,500 pull requests weekly—about 15% of all production code changes at Block. This innovation is built on Block’s open-source agent framework, Goose, and reflects the company’s collaboration with Anthropic on standards for AI model context handling, contributing to the daily use of AI across its entire engineering organization.

Block: Block is a technology company that builds financial services and developer infrastructure tools. It has focused on integrating AI capabilities into its internal engineering processes through custom agent systems. The launch of Builderbot highlights its use of these systems to accelerate software development workflows.
goose: goose is Block’s open source agent framework designed to support the building and coordination of AI agents. It serves as the foundational technology enabling tools like Builderbot for complex development tasks. The framework is publicly available to advance agent-based AI applications.
Builderbot: Builderbot is an AI tool developed by Block that orchestrates multiple agents to handle code-related tasks from within Slack. It manages research, planning, implementation, pull request creation, and iteration based on feedback. The tool operates as part of Block’s efforts to embed AI across its engineering teams.
Model Context Protocol: The Model Context Protocol is a standard co-developed by Block and Anthropic to improve how AI models handle context and tool interactions. It underpins advanced agent coordination in environments such as software development. The protocol facilitates more effective integration of AI into production workflows.

AI Integration: Block has made AI tools part of daily use for its entire engineering organization.
Industry Partnership: Block collaborated with Anthropic on standards for AI model context handling.
Open Source Framework: Block released an open source agent framework to support broader AI agent development efforts.