MCP servers tagged llm
10K live MCP servers tagged "llm", ranked by trust score. Tags come from package metadata and repository topics, so this list covers servers across every category that work with llm.
Applyos
Sistema de busqueda de empleo asistido por IA. Barre portales, analiza el calce leyendo la descripcion completa en vez del titulo, y reporta el embudo etapa por etapa.
Edith CLI
Local-first AI orchestration CLI with local LLMs, agent delegation, MCP, web research, and personal context.
Kimss Forge
Kimss Forge - MIT open-source agent harness. Any model free; gateway=kimss for production governance.
Mnemochain
Append-only, hash-chained memory and semantic recall for AI agents - local-first, citation-backed, with replay and tamper verification.
Agent Recipes
Governed procedural memory for AI agents — turn corrections, source material, and proven workflows into review-gated, versioned recipes agents can safely retrieve, lock, execute,…
Manifest Php
Manifest SDK for PHP. See every API call your app makes, heal the ones that fail, and know what will break next.
Edit2docs Web
Web studio for the edit2docs engine — generate & chat-edit PPT / Word / Excel in the browser, with live addressable-preview edit highlighting. Next.js 15 · React 19 · EN/KO.
Ocean Needle
从互联网各处捞取实用 AI Skill 的中文月刊：单个技能包、成体系的技能库，以及技能的管理 / 优化 / 安全扫描工具。每周三采集，每月 29 号出刊一期，每期最多 20 条。A monthly Chinese-language digest of practical, ready-to-use AI Skills.
Omni Io
AI customer support that degrades gracefully instead of failing. Multi-tenant RAG on NestJS, PostgreSQL RLS + pgvector, BullMQ, Gemini and React: every answer cited, traced and te…
Patchwork Harness
The verifying harness for AI coding agents: test gates, independent cross-vendor review and grounded answers, with every action on the Patchwork audit trail.
Spring Review
Offline-first code review for Spring & MyBatis: git diff in, line-level findings out. Catches @Transactional self-invocation, N+1, ${} injection, per-request thread pools.
Claude Adapter
Local Anthropic Messages API look-alike backed by the Claude Code CLI and your own subscription. Personal use; not affiliated with Anthropic.
Nano LLM Proxy
Tiny single-binary LLM gateway — pool multiple providers and API keys behind one OpenAI/Anthropic-compatible endpoint, with smart key rotation and an embedded admin GUI
Meah
Local-first MCP server that lets Claude delegate isolated work to external LLMs (Z.AI GLM, aicredits, any OpenAI-compatible API).
Early Access
Early access, feedback and Founding Hub Creator applications for Nolvion, a local-first AI environment.
Leap Framework
🎓 LEAP Framework — a state-driven agentic tutoring runtime for AI agents, exposed over MCP. Server-side State Guard, evidence-based mastery, FSRS retention.
Xiocode
Local-first terminal coding agent: self-owned runtime, kernel-backed process layer, turn-by-turn rollbacks, worktree isolation.
Troupe
A coding-agent harness for Kubernetes and your own machine: a control plane and admin console for teams, a local daemon, a terminal UI and a desktop app, all over one public proto…
Mnemosis
像人脑一样记事的 AI 记忆层：会记住、会遗忘、会整理、会自我怀疑。零依赖，支持 MCP。
N3rverberage
Internal engineering harness. Not a satellite.
Elly Server
Open-source backend for Elly -- domain logic, MCP server, and REST API for a self-hosted, ADHD-aware notebook/diary/calendar/habit/budget companion with an LLM assistant that can…
Book Agent Harness
《Agent Harness 工程》：对照 23 个真实 coding agent 的源码，讲清 agent 里模型之外那部分代码的设计决策。在线阅读 + PDF/EPUB。
Adversaria
A meeting notetaker that runs entirely on your own machine. No bot joins the call, nothing is uploaded, and the recording is deleted once it's transcribed. macOS · MIT.
Agentobs
Stop your AI coding agent before it overspends. Budget limits, guardrails and cost tracking for Claude Code - local-first, no account.
LLM Task Tree
Local-first visual task graph that keeps humans and AI agents aligned on goals, progress, reasoning, and the next action.
Wechat AI Miniprogram Guide
微信 AI 小程序开发全流程实战：多智能体 + 高德地图 MCP + FastAPI + 微信云托管。一份从 0 到真机跑通的中文指南，配一个可运行的真实项目。
Dsh Jev Verify
Jev (TypeSafe System One) for DeepSeek Harness: five verdict tools - jev_decision, jev_choose, jev_overview, jev_guard_status, jev_verify - with in-chat views, real system-prompt…
GreyWork
Desktop-first AI agent workspace for developers: ACP agents + local OpenAI-compatible models, authorized workspace with Git & file preview, plugin/MCP/skills market, scheduled tas…
Vectros Claude Code Agent Memory
Cross-session memory for Claude Code: recall relevant context automatically at the start of a turn, distill durable lessons out of a session involuntarily, and dispose them into y…
AgentDesktop
AgentDesktop - a local Windows AI agent workbench built with PySide6. 61+ tools (web search, RAG, automation, voice ASR/TTS, video generation), agent loop with sub-agent parallel…
Multimodal Desktop Agent
Modular desktop AI agent core with local/OpenAI-compatible LLMs, memory fallbacks, MCP-style tools and multimodal adapters.
Agent Reliability Toolkit
Test AI agents repeatedly, measure reliability, classify failures, and catch regressions before production.
ANNPACK
Verifiable knowledge artifacts for AI agents — package, search, and prove exactly which source revision produced each retrieval.
Agent Neutral Harness
Reliability, memory, and model-routing for AI coding agents — you keep your agent loop, pass it in as execute_fn(model, task). Local-first, MIT.
Jettison
Token optimizer for AI coding agents. Repo map, read pruning, prose compression — client-side, no proxy. 20-25% fewer turns, measured by live A/B. Verify it on your own repo.
Agent Httpd
A lightweight HTTP/1.1 server written in C — prefork event loop, CGI/FastCGI, React SSR — with a built-in native C LLM agent runtime (tool calls, MCP, skills, memory, ReAct).
Boltgo
Terminal AI agent for Ollama, OpenAI, xAI — TUI coding assistant with tools and MCP client
DelegAItor
Delegates multiple tickets — from GitHub Issues, Jira, Linear, plain markdown lists, or a Notion database — into isolated git worktrees/branches, each with its own agent session (…
Optimal Code Review Agents
Multi-agent GitHub PR reviewer (in progress): webhook → Redis fan-out → static-analysis, LLM and test-suggester agents → critic. Agents query a code graph via Graphify MCP instead…
Ctxpress
Agent 上下文压缩引擎（Java）—— 在工具输出/日志/RAG 片段进入 LLM 前做确定性、可审计、关键信息零丢失的压缩
Vyotiq Agent V
Agent V is a free, open-source coding agent for Windows, macOS and Linux. It works in your own checkout, with the model you pick, and you can roll back any edit it makes.
Repotify
Thousands of agent skills. The right ones for your repo. Repotify reads the skills developers and AI companies publish on GitHub, checks them for risks and installs the best ones…
Webgpu Agent
🤖 AI agent in your browser — WebGPU local models, skills, MCP, OpenAI/OpenRouter support
Blobfish MCP
Any OpenAPI spec. Zero config. Claude-ready. MCP server that turns any REST API into Claude-callable tools instantly.
Shop Agent
AI shopping assistant for one store. Search, cart, and Stripe checkout entirely through chat. Built with Claude, MCP, FastAPI, and pgvector, with a web storefront and a CLI.
Picoth Wire
🐈⬛ Zero pip install clutter: single-file engine turning local LLMs into Python dataclasses. Under 9MB RAM with built-in consensus voting against hallucinations.
Uriel Agent
Self-hosted family assistant: FastAPI gateway and LangGraph agent with OIDC, group-gated MCP tools and local LLMs
Pydeno
Run AI-generated JavaScript from Python, securely, in a sandbox: real V8 in a supervised, OS-sandboxed worker process (Seatbelt, Landlock, seccomp), with your Python functions as…
Tokenmarkdown
Sub-150ms Web-to-AST Markdown engine & Model Context Protocol (MCP) server. Eliminates 90%+ raw HTML token bloat for LLMs, RAG, and agent context.
Sentinel Research
A vulnerability lab and zero-trust middleware framework for Model Context Protocol (MCP). Demonstrates 8 metadata-based attack vectors and proposes a 'Sentinel' gateway for tool-p…