Packages

toolhive-catalog is a registry/catalog repository for listing Model Context Protocol (MCP) servers and reusable skills/workflows. It defines a directory structure and JSON/YAML schemas (e.g., server.json and SKILL.md) for contributors to add container-based (Docker/OCI) or remote (HTTP/HTTPS) MCP server endpoints and skill definitions.

An MCP (Model Context Protocol) server package intended to integrate “browser-use” capabilities into an MCP-compatible agent. It likely exposes browser automation/scraping tools to an LLM/agent through the MCP interface, enabling the agent to navigate web pages and perform tasks using browser control.

An MCP server package intended to expose Xenfra-related capabilities to AI agents via the Model Context Protocol (MCP).

MCP server intended to expose a “knowledge store” for use by AI agents via the Model Context Protocol. It likely provides tools for adding, retrieving, or searching knowledge/data, backed by some persistence layer.

beam_flink1.16_job_server is a Beam (Apache Beam) Flink 1.16 job server component that supports executing Beam pipelines on Flink by providing a job submission/control layer targeted to that Flink version.

docker-tika-server provides a containerized Apache Tika server setup for extracting/parsing text and metadata from documents via HTTP, using Apache Tika under the hood.

mcp-armor is a local Python CLI tool that auto-discovers and scans Model Context Protocol (MCP) configurations (for popular agent/IDE setups), inventories MCP tools/resources/prompts, runs static security checks (e.g., prompt injection, tool poisoning/shadowing, secret detection), and optionally performs baseline drift detection via an inspect/scan workflow that flags deviations from an approved baseline.

An MCP (Model Context Protocol) server package intended to integrate with Bright Data (web data/infrastructure) so an agent can perform Bright Data-related tasks through MCP tools.

An MCP server package (pointr-cloud-v8) intended to expose Pointr Cloud v8 capabilities to LLM/agent runtimes via the Model Context Protocol, enabling agents to call Pointr-related actions as MCP tools.

A Redfish mockup server that emulates Redfish (HTTP/JSON) endpoints for testing and development without a physical BMC/device.

An MCP server package intended to expose “web tool” capabilities to an AI agent via the Model Context Protocol (MCP).

An MCP server for working with Windows Remote Management (WinRM). It exposes WinRM-related capabilities to AI agents via the Model Context Protocol (MCP), allowing them to connect to Windows hosts and perform remote operations (e.g., run commands) through the MCP tool interface.

CoWork OS is a local-first desktop “AI operating system” (macOS/Windows) for running agent workflows and automations across many LLM providers and messaging channels, with a focus on security hardening, approvals for destructive actions, sandboxing/guardrails, encrypted local storage, and extensive built-in skills/connectors (including MCP connectors).

GeoServer is an open-source server for publishing and sharing geospatial data. It provides services (e.g., WMS/WFS/WCS and others depending on configuration) for clients to query and render spatial datasets from various data stores.

maestro-mcp-server appears to be an MCP (Model Context Protocol) server package intended to let AI agents access Maestro-related capabilities through MCP tools. The provided prompt does not include README/repo details, so the exact tools, parameters, and behavior cannot be verified.

MCP (Model Context Protocol) CLI server template in Rust intended as a hello-world/example implementation for MCP servers. The README describes CLI flags to enable the MCP server and optionally display prompts, resources, and tools, and shows an example configuration for integrating the server into Claude Desktop.

A Rust project that provides a collection of composable MCP servers (e.g., Discord, Shopify, Twitter/X, arXiv, Hugging Face, Replicate) that you run locally. It supports normal server mode and an SSE (Server-Sent Events) mode, configured via environment variables (e.g., Shopify domain/token).

orange-unsplash-mcp-server is an MCP server package intended to integrate with the Unsplash API to provide AI agents with access to Unsplash resources (e.g., searching or retrieving images) through MCP tools.

Provides an MCP (Model Context Protocol) server integration for TDengine (open-source time-series database), enabling AI agents to interact with TDengine via MCP tools.

Provides an MCP (Model Context Protocol) server (Java/Spring-stdio) that can search images on Unsplash, using an Unsplash access key supplied via environment variable.