Put MCP to Work
Our ToolHive project and exceptional team will help you confidently connect AI agents and models to your data and systems
Why use MCP?
Model Context Protocol (MCP) is the next big step towards an agentic future. MCP allows your AI agents to interface directly with your data and systems, so they have the context necessary to complete tasks and create real value. But MCP isn't read-only, and that means that you need guardrails.
ToolHive Makes MCP Simple and Secure
We created ToolHive as a free, open source project that developers can use to address the obstacles to MCP adoption. We're solving for identity, security, observability and more with familiar technologies like containers and Kubernetes. Enterprises using ToolHive are avoiding AI agent shelfware and using MCP to do work that matters to developers and the business.
Discover Trusted Servers
ToolHive provides a curated registry of pre-vetted servers. We remove the complexity of setup and runtime, so your team can discover and deploy with a single click.
Scale with Confidence
ToolHive runs every MCP server in an isolated container with minimal permissions. Our Kubernetes operator centralizes control of security and observability.
Protect Your Team
ToolHive keeps your secrets in encrypted storage (never in plaintext config files) and verifies software provenance to prevent malicious code execution.
Follow our insights into MCP
About Stacklok
Stacklok is founded and led by Craig McLuckie, co-creator of Kubernetes and the Cloud Native Computing Foundation. Our team of developers is applying backgrounds in security and operations to bring rigor and maturity to the fast-moving MCP landscape.