MCP servers that ship as code — a repository or a package you install and run locally (usually over stdio). They have no hosted endpoint, so we cannot probe or grade them: instead each one shows its repository, install command, license, stars and when it was last updated.
Collected from GitHub repositories and npm, PyPI, Cargo, Go, Maven, NuGet and RubyGems packages tagged as MCP servers, the official MCP Registry, Docker's MCP catalog, ModelScope, Baidu MCP World, Cline and other public directories; one entry per repository. A server that also offers a hosted endpoint is listed under Remote instead. These entries are not counted in our reachability, verification or readiness statistics.
-
ckg-nemotron-perplexitygithub.com/yarmoluk/ckg-nemotron-perplexityupdated this year
NVIDIA Nemotron + Perplexity Sonar as traversable knowledge graphs — 2 domains, 83 nodes, MCP-native agent traversal.
pypi MIT -
document-copilot-mcp-nemotrongithub.com/ADITYA22110016/document-copilot-mcp-nemotronupdated this year
RAG document assistant with a real MCP client-server architecture (Streamlit host + FastMCP server) powered by NVIDIA Nemotron.
MIT -
Your AI reviews its own code. That's the bug. An MCP server that routes code review to a different model.
MIT -
stateside-mcpgithub.com/jung-jin-lee/stateside-mcpupdated this year
MCP server for NVIDIA Nemotron-Personas-USA — 1M American personas, callable from Claude Code
NOASSERTION -
paldo-mcpgithub.com/jung-jin-lee/paldo-mcpupdated this year
MCP server for NVIDIA Nemotron-Personas-Korea — 7M Korean personas, callable from Claude Code
npm MIT -
Household operations agent: NVIDIA Nemotron on Nebius + Tavily, MCP server for Alexa+ (Amazon and Nebius hackathons)
MIT -
Opportunity triage agent: NVIDIA Nemotron on Nebius Token Factory reads listings with verified quotes; Tavily and the GitHub API check they are still open. MCP server.
MIT -
Turn one theme into a 9:16 Blender short: Nemotron Ultra plans, Super VL watches, Lightning drives Blender over MCP. (Nebius x NVIDIA Hackathon)
MIT -
Reflection-journal MCP server + web sim; Nemotron 3 Super on Nebius Token Factory with provenance chips and a deterministic crisis check. Reflection, not prediction.
MIT -
CUTROOM: a virtual cutting room. NVIDIA Nemotron agents cut your footage and build a real Shotcut timeline over MCP. Nebius x NVIDIA Global AI Hackathon.
MIT -
Private, evidence-first document agent with persistent local memory, exact citations, reversible actions, and bounded Nemotron reasoning.
MIT ★ 1 -
A personal AI that keeps your secrets and shows its receipts — NVIDIA Nemotron on Nebius Token Factory, provenance memory, privacy guard, MCP for Hermes Agent
MIT -
ai-security-gatewaygithub.com/etho0/ai-security-gatewayupdated this year
Hybrid AI Security Gateway — rule-based + LLM-assisted (Nemotron + Llama) prompt injection detection + MCP security (rules+policy+AI Assisted)
MIT ★ 2 -
Reviewable MCP security patch proposals, checked by a pinned scanner, patch audits and target tests. Offline mock rehearsal or optional Nebius/Nemotron inference.
MIT ★ 3 -
sheltergrid-backendgithub.com/Gemmaaaaaaa/sheltergrid-backendupdated this year
Agentic backend for coordinating Toronto shelter capacity — pgvector tool routing, a 31-tool FastMCP server, and an A2A mesh of Nemotron-3 Nano agents. NVIDIA Spark Hackathon 2026.
-
Voice CFO agent for small business owners: NVIDIA Nemotron on Nebius Token Factory plans, Tavily checks today's supplier prices, a deterministic cash engine says if payroll is covered. MCP server, MIT.
MIT -
Agents that design hardware, and prove it. Prompt-first workstation (Matter, Board, Silicon) on Nemotron + Nebius Token Factory and Sandboxes. Open-source pip package, CLI and MCP server. Live demo, no sign-up.
Apache-2.0 -
A private AI whose memory is a living sky: cite it, ablate it, dream on it. NVIDIA Nemotron 3 on Nebius Token Factory, with a zero-dependency MCP memory server for Hermes/Claude Code.
MIT -
nvidia-mcpgithub.com/holm-digital-io/nvidia-mcpupdated this year
MCP server that gives Claude, Cursor and any AI agent free access to 100+ NVIDIA-hosted models (Nemotron, Llama, GPT-OSS, DeepSeek, vision, embeddings) with automatic task routing and parallel multi-model fan-out.
MIT ★ 2 -
korea-persona-interviewgithub.com/binaryloader/korea-persona-interviewupdated this year
Synthetic Korean persona interview tool over NVIDIA Nemotron-Personas-Korea (CC BY 4.0). CLI / MCP server / MCP orchestrator entry points, multi-provider LLM backend (OpenAI / Anthropic / OpenAI-compatible local), multi-turn interviews with auto follow-up, persona drift detection, and prompt caching.
MIT -
GitHub Action, CLI and MCP server that replays user bug reports in a real browser under the user's own conditions, proves them with browser evidence, and opens a PR with the fix and a regression test. Runs on NVIDIA Nemotron.
Apache-2.0 ★ 2 -
Agentic security CLI that finds reachable vulnerabilities, writes the fix, and attacks its own patch in an isolated Nebius sandbox. A patch is certified only when the live exploit fails. Powered by NVIDIA Nemotron on Nebius Token Factory. Works as a CI gate and MCP server.
MIT ★ 1 -
Speaker-labeled transcripts of recorded meetings, from the command line or an AI assistant (MCP). Transcription on-device with Parakeet or Nemotron 3.5, or via Meta Muse; speakers identified on-device with Nemotron 3. macOS on Apple Silicon.
MIT ★ 1 -
persona-mcp-storegithub.com/miinhho/persona-mcp-storeupdated this year
Multi-country raw-persona MCP server over Nemotron-Personas
-
TimeWeave Memory MCP plugin for DeepSeek Harness - 4 tools, SQLite persistence, BM25 fallback, optional Nemotron embedding
-
Open-source macOS meeting transcription with GPT, FluidAudio, Nemotron alignment and MCP
MIT -
Local, private MCP server for a markdown/Obsidian vault - NVIDIA nemotron-3-embed-1b semantic search, pure-JS int8 vector store, read/write/delete tools. Local only, never published to npm.
MIT -
Records video calls (Zoom, Meet, Teams, any window), transcribes them locally with NVIDIA Nemotron, and gives AI clients the transcript and key frames over MCP. Fully local, no telemetry; it never summarizes.
MIT -
Production-grade hybrid Model Context Protocol (MCP) ecosystem featuring 4 FastMCP servers (2 Local stdio + 2 Remote SSE), custom multi-server client orchestrator, NVIDIA NIM Nemotron-3.5-Lightning LLM integration, and real-time JSON-RPC protocol inspector.
MIT ★ 1 -
Local dictation and meeting notes with Whisper or the OpenAI API. Custom Whisper models, MCP history access, and shared engines over LAN/Tailscale. Optional Parakeet/Nemotron (Windows x64, Linux x86_64), Qwen3-ASR/Moonshine (Windows x64), and Parakeet CPU/experimental MLX (Apple Silicon).
MIT ★ 269 -
MCP server that runs the Zonic agents' API, function and database tools
Apache-2.0 ★ 1 -
Autonomous Airline Disruption Recovery Agent built with NVIDIA NemoClaw, OpenShell, NeMo Retriever, and cuOpt.
-
Remote MCP server for bounded adversarial checks during AI assistant workflows, with caller-side evaluation.
MIT -
Teire from the terminal: scan a repository for the dependency and API changes that affect it, gate CI on them, and serve the same tools to coding agents over MCP.
npm MIT -
MCP server, Next.js frontend, and LangGraph chat assistant for Indian legal research (Constitution, IPC, CrPC, CPC, Evidence Act, BNS/BNSS/BSA 2023, commercial statutes, landmark Supreme Court judgments).
Apache-2.0 -
build-a-clawgithub.com/Juliusolsson05/build-a-clawupdated this year
Gothenburg open-data MCP server — federates 15+ Swedish public APIs (Västtrafik, SMHI, Trafikverket, Polisen, Styr & Ställ, …) behind one Model Context Protocol endpoint. Technical Category winner — NVIDIA Build-a-Claw 2026.