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.
-
CPU-only LLM inference engine in pure Rust — 4-bit quantized models, hand-written AVX2 kernels, speculative decoding, and a local HTTP/MCP server. No GPU, no Python runtime.
Apache-2.0 ★ 16 -
Self-hosted MCP server and Tavily-compatible API for AI agents: web search, extract, crawl and map in one lightweight CPU-only Python process. JS pages via headless Chromium, PDF, no API keys.
AGPL-3.0 ★ 3 -
Read-only MCP server for VMware ESXi: CPU Ready, disk latency, snapshots, real space usage, hardware health and events
pypi MIT -
Offline knowledge memory for AI agents: Kiwix + Wiki.js + pgvector over MCP and REST. CPU only, no GPU or AVX needed.
AGPL-3.0 ★ 1 -
Persistent local MCP/CLI memory system for AI agents. CPU-only, 98.94–99.15% R@5 on LongMemEval. Named after Mnemosyne (Greek: μνήμη, memory).
MIT ★ 2 -
Download-only BitTorrent client in Rust: ~20 MiB RAM, ~4x less CPU than Transmission, a live terminal UI and an MCP server for LLMs
MIT -
Local-first, LLM-agnostic memory system — hybrid retrieval (vector + lexical + graph), multi-signal ranking. Runs on 4GB RAM, CPU-only, no cloud.
MIT ★ 18 -
Self-hosted, source-cited RAG for OpenCode and AI coding agents. LanceDB + MiniLM hybrid retrieval, MCP server, verbatim cited answers with abstention, OmniRoute routing. CPU-only.
MIT -
Zero-dependency local translation engine via Transformers.js NLLB model — translate text across 200+ languages, CPU-only, no cloud APIs, everything stays on your machine.
npm MIT -
tfmodsearchgithub.com/santyagoseaman/tfmodsearchupdated this year
TFModSearch — MCP server for AI agents (Claude Code, Codex, Claude Desktop) to search Terraform AWS modules with hybrid search engine combining semantic similarity (sentence-transformers), BM25 text relevance, and keyword matching. FastMCP-based, CPU-only inference.
pypi MIT License -
Qdrant control plane - 3D similarity dashboard plus a read-only MCP server (9 tools). Local only, no paid API.
MIT ★ 1 -
Cache-aware context engine & state governor for AI agents: standard MCP server, lossless pruning, prefix integrity.
MIT ★ 2 -
Portable MCP context server for local and cloud LLM workflows: SQLite + FTS5, 17 tools. Early (v0.1.x), not production-ready.
npm MIT ★ 1