Packages

Provides an MCP server that connects an agent to Google BigQuery so the agent can perform BigQuery operations (e.g., query execution and related data access) via MCP tool calls. The package name indicates BigQuery-specific MCP functionality.

Provides a Go-based MongoDB-backed server that exposes gRPC operations for managing models, contexts, executing protocols, tracking protocol execution status, and storing/retrieving arbitrary data. It also mentions an MCP-style (Model Context Protocol) and Cursor IDE integration via a /mcp prefix.

NDepend.MCP.Server is an MCP (Model Context Protocol) server package intended to expose NDepend functionality to AI agents via MCP tools/handlers so agents can query and work with .NET code analysis results.

Nuitka is a Python compiler that translates Python code into generated C/C++ code and then into an executable or extension module, producing binaries that run compatibly with CPython behavior (with options for different packaging modes like standalone/onefile).

MCP server for Oracle OCI Zero Trust Packet Routing (ZPR), enabling AI agents to manage security policies that control packet routing at the network layer using zero-trust principles on Oracle Cloud Infrastructure — defining which workloads can communicate with each other based on identity rather than IP addresses.

python-ai-kit is a project template/boilerplate for generating production-oriented Python applications for AI agents (including multi-agent orchestration and state), MCP servers, and FastAPI-based microservice/monolith APIs. It emphasizes observability, testing/evaluation pipelines, prompt versioning, and (in template features) security practices like encrypting API keys.

qBittorrent is a BitTorrent client (UI + API) for managing torrent downloads/uploads. The package typically refers to running qBittorrent and interacting with its Web UI/API to add torrents, monitor status, and control sessions.

ragflow-mcp-server-aider is an MCP server wrapper for RAGFlow, intended to expose RAGFlow capabilities (e.g., retrieval/knowledge-base style operations) to an MCP-capable agent such as Aider. The name suggests it is specifically packaged for agent workflows rather than a public REST API product.

scrapy-mcp-server is an MCP server intended to help AI assistants inspect and debug Scrapy spiders and automatically generate fixes when scraping breaks (e.g., selector changes, A/B layout variants, and anti-bot changes). The README frames a workflow where monitoring/CI triggers the MCP-based debugging and produces a pull request with updated spider logic.

Search-R1 is an open-source reinforcement learning (RL) framework for training “reasoning-and-searching interleaved” LLMs. It supports RL methods (e.g., PPO/GRPO/reinforce), multiple base LLMs, and pluggable search/retrieval engines (local sparse/dense retrievers and online search). It also supports launching a separate local retrieval server that the LLM can call via an HTTP search/retrieve API during training/inference.

Senzing API Server package that exposes Senzing data/entity resolution capabilities over a network interface (providing endpoints to manage ingestion and query entities/records, backed by Senzing configuration and databases).

tika-server is a server application that exposes Apache Tika’s document parsing and text/extracted-content capabilities over a network interface, enabling clients to upload documents and receive extracted text/metadata (e.g., for indexing or content pipelines).

unirt.amplify-doc-mcp-server appears to provide an MCP server that exposes “Amplify docs” related content/operations to an AI agent. This package is intended to be used through an MCP-compatible host so the agent can query or retrieve Amplify documentation resources in a tool-call workflow.

A community MCP server package intended to be run as an MCP (Model Context Protocol) server, exposing tools to an MCP-capable agent. This evaluation is limited because no README/repo metadata or manifest content was provided to verify what tools it implements, configuration options, or operational behavior.

Provides a Zed editor MCP (Model Context Protocol) server extension that wraps the external 2slides-mcp MCP server, exposing tools to generate slide decks and to query available themes and job status/results.

Provides a Zed Editor context server/extension that exposes the ask-starknet MCP server tools inside Zed, letting an AI assistant query Starknet blockchain data and (optionally) deploy/interact with smart contracts using configured RPC and LLM provider keys.

An MCP (Model Context Protocol) server that exposes LLM-callable capabilities to track Bitcoin prices by calling the CoinGecko API (api.coingecko.com/api/v3).

DB-GPT is an open-source agentic AI data assistant that connects to data sources (databases, CSV/Excel, documents/knowledge bases), uses an LLM to generate SQL and Python/code for analysis, executes workflows (including reusable “skills”) and produces reports/charts, with an emphasis on sandboxed execution for code/tools.

gorse-server is the server component for Gorse, an open-source recommendation system. It typically provides APIs and serves recommendation results (and related indexing/training workflows) backed by a datastore.

human-skill-tree is an open-source “skill tree” that provides a curated set of AI agent skills (prompt/config templates) to transform general chat/code assistants into structured, cognitive-science-inspired learning companions (e.g., Socratic tutoring, active recall, spaced repetition). It also includes a web demo/classroom app built with Next.js + Vercel AI SDK + Supabase.