Packages

An MCP server intended to extract information from PDF documents (typically via text/content extraction and optional parsing), exposing that capability to AI agents through the Model Context Protocol (MCP).

prebid-server (PBS) is an open-source auction server for Prebid.js that provides ad server interoperability: it receives bid requests from a client, routes them to configured bidder adapters, normalizes/validates bid responses, and returns compatible bids to the publisher in the format expected by ad servers.

Provides an MCP server (MCP = Model Context Protocol) for the pydmnrules rule-engine library, enabling an AI agent to interact with DMN rules context and operations via MCP tools.

seekdb-mcp-server provides an MCP (Model Context Protocol) server that exposes Seek/SeekDB-style functionality to AI agents, enabling agents to query/search and retrieve data from a SeekDB backend via MCP tools.

An MCP server wrapper (named for a “vefaas sandbox”) intended to expose tools/resources to an MCP-capable agent environment. No repository/README contents were provided, so specific tools, endpoints, and behavior cannot be verified from this prompt.

MCP server package for integrating with the Wathq (WAT,HQ) data/source. Exposes MCP tools to an AI agent so it can query or retrieve Wathq-related information via the MCP protocol.

An MCP server package ("wecom-bot-mcp-server-remotefile") intended to expose functionality for a WeCom (WeChat Work) bot and provide remote-file related operations via the Model Context Protocol (MCP).

A Zed extension that exposes DigitalOcean's MCP server inside Zed's Agent panel by downloading the @digitalocean/mcp Node package and launching it via Node. It forwards a required DigitalOcean API token and optional configuration (enabled services, API endpoint override) so Zed can use MCP tools for DigitalOcean resources.

Agience Core is an open, extensible runtime for agentic AI centered on shared, typed, versioned “artifacts” with provenance and human-in-the-loop workflows. It is positioned as both an MCP server (exposing tools) and an MCP client (consuming vendor MCP servers), and emphasizes capability-based access via scoped API keys and delegations tracked by authority/provenance.

A lightweight Go-based MCP server intended to expose some API functionality via the Model Context Protocol (MCP) so that MCP-capable agents can call it.

ContextOS is a Python-based “context intelligence” platform that provides a unified MCP (Model Context Protocol) server and a CLI, aiming to orchestrate retrieval, memory, and reasoning layers for AI agents (e.g., cognition primitives, churn-aware retrieval routing, and index lifecycle management).

Provides an Agent Skills/MCP-oriented server plus CLI scripts to automate Douyin (TikTok China) video uploads using browser automation (Chrome) and stored login cookies, including login management and optional auto-publishing.

Provides a GitHub CLI extension that installs and runs a GitHub MCP server in stdio mode, reusing the user’s existing `gh` authentication, with a short VS Code configuration example.

A demo Spring Boot (Java) client/server architecture using MCP over SSE, wired to an OpenAI API key via Spring AI. It includes Docker Compose instructions to run the stack and configure an MCP server URL for the client.

metrics-server is a Kubernetes add-on that aggregates resource metrics (CPU/memory) from kubelets and serves them to the Kubernetes API for components like kubectl top and Horizontal Pod Autoscaler (HPA).

msedge is a Node/automation-focused package for controlling Microsoft Edge (launching the browser and driving it via automation protocols). The repository name suggests Edge specifically, but no README or manifest content was provided here to confirm supported interfaces (REST/SDK/API), auth model, or operational semantics.

Command-line and locally-run AI-agent platform with many “specialist” agents plus a REST API for invoking agents/agent chains; includes integrations for Gmail/Calendar and Outlook, a local web UI, optional voice chat, and an extensibility/plugin system. Claims “zero data” via local processing and supports LLM calls using user-provided API keys.

A collection of educational notebooks and example code for learning how to build and deploy AI agents (single-agent and multi-agent patterns) using popular Python frameworks (e.g., LangChain/LangGraph, CrewAI, AutoGen, SmolAgents) and integrating tools such as MCP, along with evaluation-oriented notebooks.

prime-mcp-server appears to be part of a Prime Intellect Python monorepo providing CLI/SDK functionality for GPU resource management, remote sandboxes, environments, pod lifecycle management, evaluations, and team context switching; authentication is handled via an API key via the CLI, and there is an MCP server component included in the repository (per manifest pythonpath), though the README content provided does not describe the MCP interface details.

MCP server package (puzzle-subway) that exposes functionality to an MCP-capable agent, likely related to puzzle/subway-style operations. No README/repo details were provided here, so the exact tools, inputs/outputs, and behavior cannot be verified from observable facts.