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. Each candidate is checked by a language model from its name, description and topics, and only actual MCP servers are listed (not clients, SDKs, lists or skills). Package metadata comes from ecosyste.ms (CC BY-SA 4.0). 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.
-
A secure database access gateway for AI Agents: Empowering LLM (Agents) with database access capabilities.
MIT ★ 6 -
Behavioral safety, vulnerability probing, and product-risk auditing platform & MCP server for Generative AI & Large Language Models.
pypi Apache-2.0 ★ 2 -
A safety layer between a language model and a broker order. MCP server: seven tools over a swappable broker backend, with the failure modes tested rather than assumed.
MIT -
Instructing rehabilitation robots with language models over the Model Context Protocol. A robot-agnostic core with a deterministic safety validator between the model and the hardware, demonstrated on a Novint Falcon under ROS 2. ETH Zurich SMS Lab semester project.
MIT ★ 1 -
MCP server for session continuity and sandbox safety in LLM development workflows. Provides "pick up exactly where you left off" + "never break the main app" functionality.
MIT -
LLM-assisted, safety-gated Postgres migrations exposed as an MCP server. The LLM proposes, the rules decide.
Apache-2.0 -
Lightweight AI safety middleware that protects humans by intercepting self-harm and criminal intent in LLM prompts. Features a 3-stage safety pipeline, MCP server for agents, and automated care responses.
pypi Apache-2.0 ★ 27 -
Constitutional-AI safety guardrails for any LLM — MCP server (npm: @creedspace/mcp-server, registry: io.github.Creed-Space/creedspace-mcp-server).
npm MIT -
binevalgithub.com/darksolitaire9-hub/binevalupdated this year
A deterministic, Rust-native binary evaluation kernel for AI safety, designed to help autonomous agents self-correct and red-team LLM infrastructure without fuzzy scores.
MIT -
sentinel-aigithub.com/MaxwellCalkin/sentinel-aiupdated this year
Real-time AI safety guardrails for LLM apps. 10 scanners: prompt injection, PII, harmful content, code vulnerabilities, obfuscation detection. Sub-ms latency. Python + TypeScript SDKs. MCP proxy. Claude Code hooks.
Apache-2.0 ★ 27 -
local-dev-mcpupdated this year
Local-first MCP server that offloads code safety, boilerplate generation, dependency evaluation, project indexing, and task delegation to local LLMs via oMLX or LM Studio.
npm MIT -
hmpo-mcp-server-practicegithub.com/iNileshW/hmpo-mcp-server-practiceupdated this year
HMPO application-status MCP server practice — FastMCP + SQLite. Parameterised read-only tools, guidance resource, JSON-lines audit log. MIT licensed. Topics: mcp, fastmcp, model-context-protocol, llm, sqlite, python, ai-safety, tutorial.
MIT -
Text2SQL data-analysis agent: three handwritten LangGraph orchestrations (deterministic pipeline / autonomous ReAct loop / Supervisor multi-agent with LLM-as-Judge), 4-layer SQL safety, MCP server+client, human-in-the-loop write path with transactional dry-run. 60-case gold benchmark.
MIT -
user-query-guardgithub.com/gowthams05/user-query-guardupdated this year
A safety gateway for validating user queries before they reach LLMs, agents, tools, or RAG pipelines.
pypi MIT -
An open-source Model Context Protocol (MCP) server for local, fast, and deterministic LLM safety checks.
-
bouncer — static compliance-controls checker (UK Online Safety Act & ICO Children's Code) as deterministic rule packs. CLI + MCP. No LLM.
npm MIT ★ 1 -
excellmgithub.com/mroshdy91/excellmupdated this year
ExceLLM - MCP server for Excel live automation with LLM safety guardrails
pypi MIT License -
Guardrails service for AI agents. Default-deny tool call evaluation with LLM safety analysis, priority-ordered decision matrix, and human-in-the-loop escalations. Session recording, behavioral analysis, MCP proxy, secret redaction, and real-time audit.
pypi NOASSERTION ★ 22 -
inmotion-mcp-servergithub.com/omertaz7/inmotion-mcp-serverupdated this year
Production MCP server (Python/FastMCP) exposing 10 safety-annotated tools over Notion, n8n & Gmail to an LLM agent layer.
MIT -
A 3-stage safety guardrail agent for LLM coding assistants (Claude Desktop, Cursor, CodeX) via MCP protocol.
MIT ★ 3 -
LLM agent trades KR/US stocks & crypto behind a deterministic safety layer — approval gates, idempotent orders, fill-evidence reconciliation. FastAPI · MCP · PostgreSQL
MIT ★ 11 -
LLM projects: RAG with citations over mine safety regulations, an agent with a custom MCP server, LoRA/DPO fine-tuning, and an LLM evaluation harness.
-
Model Context Protocol (MCP) server for LLM-optimized filesystem operations with safety controls. For use with Claude Desktop.
npm MIT -
Safety-first MCP server connecting LLM agents to the Binance Spot API (testnet-first, read-only tools, guardrails).
-
ifc-consolegithub.com/nbharathik/ifc-consoleupdated this year
A terminal interface to connect IFC files to LLMs: a standalone MCP server for IFC/BIM with an ask/edit safety switch and a local 3D viewer.
pypi Apache-2.0 -
nexus-coregithub.com/Noumenon-ai/nexus-coreupdated this year
Self-hostable multi-MCP personal-assistant framework: multi-provider LLM orchestration, an agentic tool loop, an approval-gated safety model, conversational recovery, and 10 MCP servers. MIT.
MIT -
zero-ai-gatewaygithub.com/vishtechie07/zero-ai-gatewayupdated this year
Zero-Trust AI Gateway: Dual-pass PII redaction (Regex + LLM), real-time security monitoring with Kafka/WebSockets, and MCP-powered compliance auditing for enterprise AI safety.
MIT -
mcp-clickhousegithub.com/TopWent/mcp-clickhouseupdated this year
Model Context Protocol server exposing ClickHouse to LLM clients over stdio. Read-only by default, comment-aware SQL safety validator, five tools, distroless image.
MIT -
Project-Heimdallgithub.com/AnubhavMadhav/Project-Heimdallupdated this year
A high-security Model Context Protocol (MCP) server for PostgreSQL. Acts as a read-only gatekeeper for LLMs by using AST parsing to strictly enforce safety and prevent data mutation.
MIT -
Metatrader5-CLIgithub.com/ehukaimedia/Metatrader5-CLIupdated this year
Agent-native Python library + CLI to control MetaTrader 5 — uniform JSON envelopes, a safety-gated order surface, and an MCP server for LLM agents. Tool-only.
MIT ★ 1 -
commandergithub.com/drhalto/commanderupdated this year
Local control for an EZ-Robot JD humanoid, replacing Synthiam ARC - web GUI, safety guardrails, and LLM agents as the robot's operator over MCP. No account, no subscription, no cloud.
MIT -
Multi-Agent Autonomous Cybersecurity Platform (Go + LLM) — layered agent orchestration, 91 YAML tool recipes + 50+ built-in MCP tools, HITL safety, single-binary offline deployment.
MIT -
gmail-mcp-agentgithub.com/shariq-beg/gmail-mcp-agentupdated this year
Privacy-first multi-agent Gmail assistant with human-in-the-loop safety on every destructive action — built on a custom MCP server, LangGraph orchestration, and a fully local LLM.
★ 1 -
🛡️ Rust CLI extension for Claude Code — LLM-powered command validation, SQLite context persistence & MCP tools. Two-layer safety (rules + Ollama), vector semantic search, and user-adaptive learning.
MPL-2.0 ★ 1 -
sqlite-explorer-fastmcp-mcp-servergithub.com/hannesrudolph/sqlite-explorer-fastmcp-mcp-serverstale
An MCP server that provides safe, read-only access to SQLite databases through Model Context Protocol (MCP). This server is built with the FastMCP framework, which enables LLMs to explore and query SQLite databases with built-in safety features and query validation.
★ 108 -
MCP server that gives local LLMs (LM Studio, Ollama) live web access plus recon, OSINT, and authorized security-testing tools like Burp/ZAP/Wapiti integration, APK triage, and self-extending skills, all safety-gated by default.
MIT ★ 3 -
MCP server for AI-driven penetration testing. Wraps 22 security tools (nmap, nuclei, sqlmap, Metasploit…) as 46 safety-hardened MCP tools for Claude and LLM agents. Written in Rust.
oci Apache-2.0 ★ 3 -
local-llm-mcpgithub.com/HenryLinyy/local-llm-mcpupdated this year
MCP server that lets Claude Code & Codex delegate bounded, token-heavy generation to local or cheap-cloud LLMs — LM Studio · Ollama · vLLM · llama.cpp · DS4, plus DeepSeek/OpenRouter/Groq/Cerebras/MiniMax/Agnes. Keep the frontier agent's judgment, cut the token bill, with RAM guards and a text-only safety boundary.
MIT ★ 3 -
ButtplugLLMgithub.com/zhanp199/ButtplugLLMupdated this year
Chat with a local LLM that controls intimate hardware via the Buttplug.io protocol — FastMCP device server, LM Studio orchestration, multi-chat web UI with vibration pattern templates and built-in safety (e-stop, safe-word, watchdog).
BSD-3-Clause ★ 7 -
sql-explorer-mcpgithub.com/pawansingh3889/sql-explorer-mcpupdated this year
Read-only Model Context Protocol server for SQL databases. Lets LLMs (Claude, Cursor, ChatGPT, Continue) introspect and query SQL Server, Postgres, and SQLite with multi-layer safety guards including sql-sop linting.
pypi MIT -
MCP server exposing PubMed, ClinicalTrials.gov, openFDA, and ChEMBL as LLM-callable tools for biomedical research, trial discovery, drug-safety analysis, and innovation forecasting.
-
Intent-first search that finds the right tool from a vague description. Hybrid retrieval over LLM-written cards, LLM rerank with recall safety, MCP server for AI agents, runs on a laptop CPU.
MIT ★ 1 -
MCP server wrapper for EvilLimiter. Provides LLM-accessible tools for network bandwidth control, monitoring, and traffic analysis. Docker-based with safety features.
MIT ★ 7 -
A production-grade, self-hostable MCP server for Quera, bringing Quera’s problemset, submissions, courses, contests, LMS, AI, judge tooling, and authoring workflows to LLM agents. Built with async Python, 259+ tools, RBAC, safety gates, audit logs, revisions, multi-account support, and Docker deployment.
MIT -
operandi-mcpupdated this year
MCP server — let any LLM agent identify an appliance and pull its grounded, safety-checked, robot-executable operation package from OPERANDI.
pypi Proprietary -
deepl-i18ngithub.com/lancedelgardo/deepl-i18nupdated this year
DeepL-first translation of localization files (JSON, YAML, .arb, .strings, .properties, Android XML, CSV, gettext), with placeholder-safety, a web UI, an MCP server and a local-LLM fallback.
pypi MIT -
🧠 Save 70–95% tokens in Claude Code, Cursor, Windsurf & Cline. Local Tree-sitter code graph MCP server: surgical edit context instead of full-file reads, refactor safety checks, and a hook that blocks large file reads. Reduce token usage & cost, zero LLM indexing.
pypi MIT ★ 4 -
Static-analysis MRI for codebases — a Tree-Sitter dossier engine + deterministic PreToolUse safety gate for AI coding agents. Zero LLM calls, zero API keys.
MIT -
MCP injection / prompt-poisoning / SSRF scanner. 30+ canonical rules covering the April 2026 Anthropic MCP RCE class, mcp-server-git CVE chain, DockerDash injection chain, OWASP LLM Top 10 + GenAI Red Team v1. Scans live MCP /tools/list endpoints OR pasted JSON. Free 5/day, Pro £29/mo for unlimited + signed safety reports auditors and procurement accept. By MEOK AI Labs.
pypi MIT License -
This MCP server extracts key-value pairs from arbitrary, noisy, or unstructured text using LLMs (GPT-4.1-mini) and pydantic-ai. It keeps type safety and supports multiple output formats (JSON, YAML, TOML). The server is robust to any input and always attempts to structure data as much as possible.
GPL-3.0 ★ 1 -
MCP server for true402 — every true402 stall as a pay-per-call tool (token rug/honeypot safety, web, SEO, markets, LLM chat), auto-discovered from the live catalog. x402 USDC on Base; no accounts, no API keys.
npm MIT -
MCP server for IP-Symcon — lets an LLM/agent inspect, develop and control an IP-Symcon home-automation system over its JSON-RPC API: read the object tree, read/edit/create PHP scripts, read variables, control devices. Write/dev tools gated behind a safety flag.
MIT -
ipaship-auditgithub.com/atharvnaik1/ipaship-auditupdated this year
Let your AI Agent review your iOS/Android apps for appstore policy & security bugs with ipaShip's web app, cli, mcp or claude-skill with safety hooks layer any any llm agent you use - loopin fixes in dev stage only no more waits till submissions.
MIT ★ 102 -
Cross-vendor multi-model review committee that gates AI/agent-generated content before it ships — consensus-based hallucination defense. Zero dependencies, local-first CLI, MCP server, CI gate exit codes.
GPL-3.0 ★ 1 -
quorumgithub.com/marklynd/quorumupdated this year
A council of language models that scores a claim against a rubric, using only evidence you supply. Bounded timeouts, anonymised deliberation, hash-verified transcripts, MCP server.
MIT -
claude-india-dpigithub.com/romichandra/claude-india-dpiupdated this year
An enterprise-grade AI gateway framework designed to safely and deterministically scale frontier language models across India's Digital Public Infrastructure (DPI) registries via MCP, Pydantic type-guards, and SymPy validation.
-
sqlgategithub.com/asd3358123/sqlgateupdated this year
A deterministic, fail-closed write gate for LLM-generated SQL, exposed over MCP. The model can't talk its way past it.
MIT -
A safety-aware MCP server and Claude skills suit for ROS 2 Engineering
MIT ★ 4 -
MCP Server for Swiss Federal Food Safety and Veterinary Office (BLV) open data — no auth required
pypi MIT -
Reality-check layer for AI agents. Detect hidden risks, missing context & real-world consequences before your LLM answers or acts. MCP tool for Claude Code & Cursor.
NOASSERTION