Packages

MCP server that provides Solana blockchain data access to AI agents via the Model Context Protocol, allowing tools/queries to retrieve Solana on-chain information (e.g., addresses, transactions, blocks, program/asset-related data).

Toolathlon-GYM is a self-contained, locally runnable evaluation/training environment for LLM agents’ real-world tool use. It provides 503 automated multi-step tasks backed by a local PostgreSQL database and orchestrated via 25 MCP servers, running each task inside an ephemeral Docker container with automated preprocessing and evaluation scripts.

Provides an MCP (Model Context Protocol) server that lets an AI agent interact with Gitea (repositories, issues, pull requests, etc.) via MCP tools, so the agent can perform common repository management and information retrieval tasks.

Home Assistant is an open-source home automation platform that integrates with many smart-home devices and services, providing automations, dashboards, and a central control hub running locally (or via supported hosting).

mcp-client-gen generates type-safe TypeScript clients from MCP (Model Context Protocol) server schemas, including a helper for establishing an MCP connection and handling OAuth 2.1 (PKCE) browser-based auth for OAuth-protected MCP servers.

mcp-debug is a Go command-line tool for debugging Model Context Protocol (MCP) servers. It can connect to MCP servers (including via streamable-http), provide an interactive REPL to inspect tools/resources/prompts, log JSON-RPC traffic verbosely, and run in an MCP server mode itself. It also supports OAuth 2.1 authentication flow options (including PKCE, discovery, resource indicators, and optional CIMD/DRC) to access protected MCP endpoints.

Provides an MCP server (Spring AI MCP Server, stdio mode) that lets an AI tool create article drafts and publish articles to Juejin using a user-provided Juejin login cookie. Accepts Markdown content and tags, and exposes a tool method (saveArticleToJueJin) that calls Juejin content APIs.

A Spring Boot-based MCP server that exposes tools for interacting with a Nacos cluster: listing and inspecting namespaces, configurations (including history), services (and instances), and cluster/node health and metadata via an MCP SSE endpoint.

An MCP server (Python) that turns text prompts (and optionally images) into multi-view generated images, lets users approve/deny those images, reconstructs a 3D model using CUDA Multi-View Stereo (locally or via a remote GPU server on a LAN), and generates/refines/export parametric OpenSCAD-based outputs. Includes optional printer discovery/printing and a local web UI for the workflow.

MCP server exposing Oracle OCI Events service API, enabling AI agents to manage event rules, subscriptions, and notifications within Oracle Cloud Infrastructure. Allows creating, listing, updating, and deleting event rules that trigger actions when OCI resource state changes occur.

MCP server for Oracle OCI Functions as a Service (FaaS) - enabling AI agents to manage serverless function applications, deploy functions, invoke functions, and manage function configurations within Oracle Cloud Infrastructure.

MCP server exposing Oracle OCI File Storage Service APIs, enabling AI agents to manage NFS-based shared file systems including creating/deleting file systems, mount targets, exports, and snapshots within Oracle Cloud Infrastructure.

MCP server exposing Oracle OCI Functions service APIs for serverless compute, enabling AI agents to manage function applications, deploy and invoke functions, manage configurations, and handle function lifecycle within Oracle Cloud Infrastructure.

MCP server for Oracle OCI Generic Artifacts and Content service, enabling AI agents to manage artifact repositories, store and retrieve versioned artifacts, and manage artifact content within Oracle Cloud Infrastructure.

Provides a collection of MCP (Model Context Protocol) server implementations that wrap many common cybersecurity/pentesting tools (e.g., nmap, nuclei, sqlmap, zap, trivy, etc.) so they can be executed via an MCP-compatible client, including a “unified” server to access multiple tools through a single MCP connection.

A Spring Boot 3 wallet management backend (CRUD wallets plus income/expense operations) that exposes service methods as AI-accessible tools via Spring AI’s Model Context Protocol (MCP) server integration, backed by an H2 in-memory database for testing.

UTCP-MCP Bridge provides an MCP server that exposes UTCP “manuals” as MCP tools. It can be run via npx (no install step for the user) and is configured through a .utcp_config.json to register HTTP/OpenAPI-based tools, post-process tool outputs, and manage tool discovery/execution from an MCP client (e.g., Claude Desktop).

Chrome DevTools extension (and MCP tool server) that detects CSS declarations that have no visible effect on the selected DOM element or across the full page. It uses computed styles plus heuristics, and reports actionable warnings with fix suggestions. The MCP server exposes the rules engine for AI-assisted analysis using Playwright.

DevDocs is a self-hostable documentation crawler and extraction tool that can crawl developer documentation sites (with depth control and link discovery), extract cleaned content, and expose it via an MCP server for LLM querying. It also provides a local UI and backend services via Docker, and supports exporting crawl results (e.g., MD/JSON) for further use.

DocSentinel is a Python/FastAPI MCP-ready service that parses security documents (PDF/DOCX/XLSX/PPTX/text), indexes an organization’s security policies into a knowledge base (RAG), and uses configurable LLM backends to generate structured security assessment reports (risks, compliance gaps, and remediation suggestions). It exposes REST endpoints for assessments and knowledge-base operations and includes an MCP server for agent integration.