Bolt

•
372
•

Orchestrates, deploys, and manages enterprise-grade custom AI-powered agent servers on-premises.

Bolt is an enterprise-grade orchestration platform designed for creating, deploying, and managing custom AI agent servers within on-premises environments. Evolving from the OpenMCP project, Bolt focuses on providing secure, self-hosted solutions with robust cloud-native infrastructure integrations. It offers rich customization capabilities for tools, allowing for declarative input/output transformation and composition of multiple API operations. With features like Streamable HTTP and OAuth 2 authorization for all servers, a bundled client playground, an official Inspector, and the flexibility to integrate your preferred LLM, Bolt empowers users to build and manage tailored AI agent solutions effectively.

Key Features

01Rich customization of tools, supporting declarative JSON/YAML specs for input/output transformation and API composition.
02Flexible LLM integration, allowing configuration of self-hosted or cloud-based large language models.
03Secure Streamable HTTP communication with OAuth 2 authorization by default for all servers.
04333 GitHub stars
05Self-hosted on-premises deployment with enterprise-grade cloud-native infrastructure.
06Bundled MCP client playground and official Inspector for interactive testing and server interaction.

Use Cases

01Deploying and managing custom AI agent servers in secure, on-premises enterprise environments.
02Creating tailored agent servers with specific subsets of tools for specialized business or operational needs.
03Orchestrating AI agents that interact with multiple upstream APIs, each with unique authentication requirements.