📄️ MCP primer
Model Context Protocol (MCP) connects AI assistants to external tools and data sources through a standard interface.
📄️ Organizing MCP servers
Groups organize MCP servers into logical sets and control which clients can access them.
📄️ Optimizing LLM context
Tool filtering and overrides reduce context window pollution and improve AI tool selection accuracy.
📄️ Registry criteria
Requirements and review criteria for adding MCP servers to the ToolHive registry.
📄️ Observability
Observability features in ToolHive, including metrics, traces, and logging for MCP server monitoring.
📄️ Authentication and authorization
ToolHive's authentication and authorization framework for securing access to MCP servers.
📄️ Cedar policies
Writing and configuring Cedar policies for MCP server authorization.
📄️ Backend authentication
Backend authentication patterns for MCP servers, including static credentials, token exchange, and the embedded authorization server.
📄️ Embedded authorization server
How the ToolHive embedded authorization server works, including the OAuth flow, token storage and forwarding, and when to use it.
📄️ Virtual MCP Server (vMCP)
Virtual MCP Server (vMCP) aggregates multiple MCP servers into one endpoint for simpler client configuration.
📄️ Understanding skills
Skills are reusable MCP server configurations that bundle tools, permissions, and secrets for common tasks.