Packages

A TypeScript framework for building AI agents that can interact with Model Context Protocol (MCP) servers and local tool functions. It provides agent abstraction, an MCP connection manager for reusing server connections, per-agent MCP tool aggregation, LLM provider integration (example for Fireworks AI), simple in-memory message history, and example orchestration/workflows.

Perplexity-focused MCP server that exposes tools to an MCP-capable agent so it can perform Perplexity-powered search/browsing and return results in the agent workflow.

robotmcp-server is a modular Model Context Protocol (MCP) server (Python/FastAPI) that discovers/loads MCP tools from git submodules, exposes an MCP-compatible HTTP transport endpoint, and can secure access via OAuth 2.1 with Supabase-backed user/session handling. It also supports Cloudflare Tunnel access and provides CLI commands for module management and connectivity verification.

Provides a local MCP (Model Context Protocol) server that wraps the CData JDBC Driver for Sage Cloud Accounting, exposing the underlying data as MCP tools (get tables, get columns, run SQL SELECT) so an LLM client can query live accounting data without writing SQL manually.

Provides a read-only Model Context Protocol (MCP) server that exposes Salesforce Data Cloud as relational SQL tables by wrapping CData's JDBC Driver. It runs as a local process using stdio and offers MCP tools for listing tables/columns and running SQL SELECT queries.

Provides a local, read-only MCP server wrapper around the CData JDBC Driver for SAP SuccessFactors. It exposes SuccessFactors data as SQL tables/models and lets an MCP-capable client (e.g., Claude Desktop) discover tables/columns and run read queries via MCP tools over stdio on the same machine.

A Spring Boot–based MCP server that integrates with Tencent Cloud Developer Community, exposing an MCP tool (addArticle) to publish Markdown articles. It converts Markdown to ProseMirror-like content for compatibility with Tencent’s editor and returns status/URLs/error messages. Authentication is done via a user cookie configured in application settings or environment variables.

Provides an MCP server exposing UCP (Universal Commerce Protocol) shopping flows as callable tools for AI assistants, including merchant capability discovery, checkout creation, checkout updates (discounts), fulfillment/shipping setup, and checkout completion/payment submission.

Provides an MCP (Model Context Protocol) server that integrates an AI assistant with a YAPI API management instance. It exposes MCP tools to list YAPI projects/categories/interfaces, search interfaces, fetch interface details, and manage/refresh a local cache (including cache stats and per-project cache clearing).

Provides a Model Context Protocol (MCP) server (Go) that wraps the Etherscan API V2 to let LLM applications query multi-chain blockchain data such as balances, blocks, transactions, contract ABI/source, gas oracle, and token/NFT transfers.

Provides an MCP server (implemented in Go) that converts provided SVG input into favicon assets in ICO and/or PNG formats (e.g., 16x16, 32x32, 48x48). Output is returned as base64 in MCP responses and can optionally be written to files on disk.

Provides an MCP server that uses browser automation to search LinkedIn for people, companies, and jobs, scrape profile/feed data, and return structured JSON via MCP tools (stdio or streamable-http).

Provides an MCP server that integrates MATLAB with an AI assistant. It exposes MCP tools to (1) execute MATLAB code and (2) generate MATLAB code from natural language, and it exposes a resource/URI to access MATLAB documentation.

Provides a .NET Model Context Protocol (MCP) App with an interactive HTML color picker UI. Exposes an HTTP MCP endpoint at /mcp and a UI/resource endpoint for rendering the picker, with a tool named ColorPicker that accepts an optional initialColor (hex) and returns a result containing the UI resource URI to display.

mcp-chat is an open-source chat application template/demo that integrates Pipedream’s MCP server to enable AI chat-based tool calling across many third-party APIs, using the Vercel AI SDK for model/tool orchestration and optional persistence/sign-in.

mcp-kit is a Go-based toolkit that provides an MCP server and MCP client implementation (using github.com/shaharia-lab/goai/mcp), plus an HTTP API server layer that exposes REST endpoints and orchestrates requests between a frontend, the MCP client, and LLM providers. It also includes monitoring/observability via Prometheus/Grafana/Loki/Promtail and provides an OpenAPI schema file (openapi.yaml) for the API server.

A TypeScript bridge that connects Ollama-hosted local LLMs to MCP (Model Context Protocol) servers via JSON-RPC. It registers MCP tools, routes tool calls produced by the LLM to the appropriate MCP server(s), and validates/returns structured results to the user.

SQL Server Performance Monitor provides free, open-source performance monitoring and alerting for SQL Server and supported cloud SQL platforms. It collects performance data via scheduled T-SQL collectors (Full edition uses SQL Agent and installs a monitoring database; Lite edition runs a desktop app and stores data locally using DuckDB/Parquet), renders dashboards and plan visualizations, and includes a built-in MCP server offering read-only tools for AI-assisted analysis based on the collected monitoring data.

Provides an MCP server that exposes OCR capabilities (based on RapidOCR) via two tools: OCR from base64-encoded image content and OCR from an image file path.

**This repository is deprecated and no longer maintained.**