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.
-
Stop RAG hallucinations: verify every LLM answer against its sources and say "I don't know" instead of guessing. TypeScript, Python, MCP server.
npm Apache-2.0 -
Python SDK for Komplex AI — a hallucination detector for LLM output. Score any response for hallucination risk in one call. Includes an MCP server.
NOASSERTION -
haluguard-mcpgithub.com/haluguard/haluguard-mcpupdated this year
MCP server for haluguard — hallucination detection for LLM apps. Plugs into Claude Code, Cursor, ChatGPT Desktop and any MCP-compatible host.
npm MIT -
grievous-mcpgithub.com/darthcoder/grievousupdated this year
A hallucination engine. Typed pseudorandom data via LLM. No guarantees. Trigger discipline required.
pypi MIT -
mcp-oracologithub.com/simoneloru/oracoloupdated this year
MCP (Model Context Protocol) server designed to solve LLM code hallucinations by performing local static analysis.
npm ISC -
MCP server providing LLMs direct access to local package source code to eliminate API hallucinations
npm MIT ★ 1 -
mcp-knowledge-graphgithub.com/roberteisenberg/mcp-knowledge-graphupdated this year
Clinical intelligence tool built across 6 phases — demonstrates reducing LLM hallucinations and cost by moving work into infrastructure
★ 2 -
OptiMCPgithub.com/vogue84/OptiMCPupdated this year
Stop shipping hallucinated numbers. OptiMCP verifies your agent's math and logic against your rules — deterministically, with zero LLM in the loop — and names the exact rule that broke.
NOASSERTION ★ 6 -
Judges agent/RAG outputs (relevance, hallucination, accuracy, routing, prompt injection) using Laya instead of LLM-as-judge -- usable as an MCP server or as a plain Python library.
pypi MIT -
couchloop-eq-mcpgithub.com/wisenbergg/couchloop-mcpupdated this year
Behavioral governance layer for LLMs - monitors for hallucination, inconsistency, tone drift, and unsafe reasoning while managing stateful AI sessions
npm MIT -
Human-in-the-loop LLM Failure Detection & Benchmark Toolkit — Auto-flag LLM failures (hallucination, refusal, factual error) with LLM-as-Judge explanations. Confirm them in a review UI or via MCP agent — then export a citable benchmark validated against expert labels. Compatible with Ollama and Amazon Bedrock models.
-
hallumarkgithub.com/cognis-digital/hallumarkupdated this year
LLM hallucination & grounding auditor for RAG systems
NOASSERTION -
us-law-mcpgithub.com/seelpeed-debug/us-law-mcpupdated this year
MCP server for U.S. federal law — search the U.S. Code, CFR, Federal Register, case law and bills, and verify citations to stop LLM hallucinations
MIT -
.NET SQL Agent MCP server featuring raw SQL input, strict AST validation, and an embedded Admin UI. Eliminates LLM hallucinations and security risks across 6 major databases.
nuget Apache-2.0 ★ 11 -
lingjing-classroom-mcpactive
LingJing Classroom — a deterministic (no-LLM) knowledge & blindspot MCP for AI agents: a teaching loop that returns 7-dimension blindspots and a summary. An independent gap-check layer, never hallucinates, never grades.
npm UNLICENSED -
> **Model Context Protocol (MCP) Server for ColabAI** > Direct academic citation verification, anti-hallucination paper validation, and multi-source academic search for AI agents and LLM tools (Cursor, Claude Desktop, Windsurf, LangChain, OMP).
npm MIT -
openclaw-consensus-mcpgithub.com/yanmiayn/openclaw-consensus-mcpupdated this year
MCP server for OpenClaw Consensus ??9-LLM consensus to fight hallucinations. Built by yanmiayn, 16yo Korean solo dev.
pypi MIT -
doloop-machinegithub.com/ekras-doloop/doloop-machineupdated this year
Distributed Collection, Local Intelligence - Stop LLMs from hallucinating with Kong in the Loop architecture
pypi MIT ★ 1 -
CodeShield-MCPgithub.com/Thejosem4/CodeShield-MCPupdated this year
MCP server and CLI that prevents LLM hallucinations by verifying code against known libraries and frameworks
npm MIT -
Self-hosted multi-agent AI platform: generates, evaluates and refines LLM answers until they hit a quality bar (~22% more accurate, ~45% fewer hallucinations)
MIT -
ai-context-mcpgithub.com/prathmeshkunturwar/context_mcpupdated this year
MCP server that gives LLMs live, version-pinned documentation from 57+ AI/ML libraries. Eliminates hallucinations about APIs.
npm MIT -
metricmate-mcpgithub.com/marilynceo/metricmate-mcpupdated this year
LLM Output Quality Evaluation u2014 quantify, evaluate, and track LLM response quality: relevance, accuracy, tone, hallucination scoring, and consistency checks
-
insight-link-progithub.com/Lonishubh48/insight-link-proupdated this year
Hallucination Killer — MCP server that grounds LLM answers in live repo context and real-time docs
★ 2 -
dotnetaimcpservergithub.com/arbelh/dotnet-ai-mcp-serverupdated this year
MCP Server for .NET AI development - provides real code samples from Semantic Kernel, OpenAI .NET, MCP C# SDK, and more. Prevents LLM hallucinations by fetching live code from GitHub repositories.
nuget MIT -
PubMed-MCPgithub.com/Mostafao4/PubMed-MCPupdated this year
Model Context Protocol server providing verifiable PubMed research data so LLMs can reason safely without hallucinating biomedical facts.(Hakathon)
★ 1 -
A deterministic first-pass filter for unsupported claims in LLM output, computed without calling another LLM. Not a standalone hallucination detector. A spec plus reference implementation, shipped as touchstone-mcp.
pypi Apache-2.0 -
rag-eval-enginegithub.com/aptsalt/rag-eval-engineupdated this year
Production RAG with continuous evaluation — hybrid retrieval (BM25 + vector/RRF), LLM-as-judge scoring on faithfulness/relevance/hallucination, MCP server, adaptive retrieval tuning. 68 tests.
MIT -
atlasgithub.com/2005-Aneeshdutt/atlasupdated this year
Automated LLM red-teaming pipeline: a LangGraph agent runs jailbreak, prompt-injection, hallucination and bias attacks against a model, scores responses with an LLM judge, mutates failed attacks, and reports results in a Streamlit dashboard and MCP agent.
MIT ★ 1 -
.NET MCP server that gives LLMs real NuGet API metadata to reduce hallucinations
nuget NOASSERTION ★ 24 -
A simple MCP server that provides conversion between hexadecimal and decimal numbers. Mainly solving the problem of LLM hallucinations.
MIT -
secure-mcp-auditorgithub.com/dhzkh68cgj-cmyk/secure-mcp-auditorupdated this year
A verifiable audit layer for LLM tool-calling using Model Context Protocol (MCP) to detect hallucinations and prevent logic sabotage.
-
Locker is a lightweight MCP that enforces file-level access control to reduce LLM Loops, hallucinations and inconsistancies.
npm -
The Karpathy LLM Wiki, production-ready. Zero-RAG personal knowledge base with MCP server, multi-agent support, knowledge graph visualization, and hallucination enforcement. No embeddings, no vector DBs — just markdown + git.
★ 5 -
A RAG agent that grades its own retrieval, reformulates the query when confidence is low, and abstains instead of hallucinating. LangGraph + hybrid search + reranking, served over MCP, measured against two naive-RAG baselines with an LLM-as-judge harness and RAGAS.
MIT -
ReasonForgegithub.com/foxintheloop/ReasonForgeupdated this year
Modular Mathematics Toolkit MCP Server: 111 tools that give LLMs exact math instead of hallucinated guesses. Routes calculations through SymPy/NumPy/SciPy for provably correct results. 7 modular packages: logic, algebra, calculus, statistics, physics. Load only what you need.
pypi MIT -
biocite-mcpgithub.com/muslus/biocite-mcpupdated this year
🧬 Verified academic citations & literature discovery for LLMs. Built with Europe PMC & Crossref. No more hallucinations! 🚀 Advanced MCP server for verified biological literature discovery, professional BibTeX export, and manuscript reference auditing.
MIT -
Model Context Protocol (MCP) server that gives LLM agents and chatbots direct access to the World Bank's Data360 Platform. Search, validate, and retrieve development indicators—from GDP and poverty to gender equality and climate—with structured metadata and time-series data, without hallucinating values.
go NOASSERTION ★ 45 -
Local, version-exact API index of the packages installed in your Python/TS projects, with a static checker that catches hallucinated or misused APIs in code an LLM just wrote - without ever running that code. FastAPI + React, MCP server included.
MIT -
PR-SENTINELgithub.com/Kapish7717/PR-SENTINELupdated this year
An autonomous GitHub App review agent built with LangGraph, FastMCP, and Groq (LLaMA 3.3). Combines deterministic static analysis (Ruff) with advanced LLM reasoning to catch bugs, styling nits, and security risks—validated with evidence-checking to prevent hallucinations. Runs asynchronously via webhook integration.
-
ejentum-mcpno update data
Exposes the four Ejentum cognitive harnesses (reasoning, code, anti-deception, memory) as MCP tools any agentic client can call. Drop-in scaffolding that catches LLM failure modes like sycophancy, hallucination, and reasoning shortcuts.
smithery -
fusion-mcp-hubgithub.com/KirtiJha/fusion-mcp-hubupdated this year
Solve the MCP token bloat problem. Build lean, custom MCP servers with only the tools you need from 300+ available tools. Reduce LLM costs and hallucinations. Includes CLI, WebSocket proxy, AI-powered recommendations, and interactive playground.
MIT ★ 2 -
An MCP Server for APIs.guru remarkable Wikipedia for Web APIs. Providing tools, prompts and resources to help eliminate LLM Coding Agent hallucinations.
npm MIT ★ 9 -
weather-mcp-a2agithub.com/Ratnesh-181998/weather-mcp-a2aupdated this year
MCP-based agentic AI system enabling LLMs to fetch and reason over real-time global weather data using standardized tool invocation. Built with MCP Server/Client, Llama 3, and external APIs to deliver low-latency, hallucination-free weather intelligence through protocol-driven workflows.
MIT ★ 1 -
🚀 Supercharge your AI agents! A cross-platform CLI & Model Context Protocol (MCP) server that parses full codebases into token-efficient, LLM-ready context files. Save tokens, reduce hallucinations, and give your AI the perfect context window.
MIT ★ 1 -
MCP server for the Pionex exchange with strict anti-hallucination guardrails: two-phase order confirmation, live symbol/price validation, operator-set limits. Works with Claude and local LLMs.
MIT -
calculus-helper-mcp-demogithub.com/rdwj/calculus-helper-mcp-demoupdated this year
MCP server that offloads symbolic calculus to SymPy — 8 tools (derivatives, integrals, limits, Taylor series, equation & ODE solving, simplification, numeric evaluation) so LLM agents get computed answers instead of hallucinations. Built on FastMCP 3.x.
NOASSERTION -
Persistent visual cache for LLM-driven software development. Caches screenshots using perceptual hashing, vector search, and AX trees to prevent token overhead and visual hallucination loops.
npm NOASSERTION ★ 41 -
Magneto is a non-LLM adversarial citation gate whose job is to prevent hallucinations by fact-checking an agent in real-time. Currently for Kiro. EXPERIMENTAL.
go NOASSERTION -
An MCP server that bridges Anthropic’s Claude (and other LLMs) with the CourtListener API for live legal data—opinions, dockets, court records, and more. Designed to power legal research and compliance tools with transparency, reliability, and fewer hallucinations. Scalable, open, and easy to extend.
npm MIT ★ 16 -
zadeh.aigithub.com/deegitech/zadeh-netupdated this year
The AI companion to Zadeh.NET: turn LLM confidence into deterministic Execute/Confirm/Clarify/Reject decisions (ConfidenceGate), generate fuzzy rule sets from plain-language policy via any LLM (RuleSmith), compute adaptive similarity thresholds for RAG/semantic caching (AdaptiveThreshold), and expose fuzzy engines as MCP tools for AI agents. AI designs, fuzzy decides — zero hallucination at runtime. Uses only the .NET base library — no external dependencies. Licensed under AGPL-3.0 / Commercial dual license.
nuget AGPL-3.0-only -
A deterministic MCP server and skill for OSHA recordkeeping (29 CFR Part 1904). Evaluates workplace injury recordability via 11 typed triage tools and 15 bundled eCFR datasets. Eliminates LLM hallucination in compliance lookups with auditable citations, zero data retention, and support for stdio, Node HTTP, and Cloudflare Workers.
MIT ★ 1 -
nexus-mcp-extensiongithub.com/tanaikech/nexus-mcp-extensionupdated this year
Nexus-MCP solves Tool Space Interference in LLMs. It uses a single gateway to aggregate multiple MCP servers and employs a deterministic 4-phase workflow (Discovery, Mapping, Schema Verification, Bridged Execution) to prevent context saturation and tool hallucinations, maintaining reasoning accuracy with massive tool ecosystems.
MIT ★ 4 -
Autonomous DFIR triage agent built with Google ADK & FastMCP. 4-phase NLP pipeline: frequency stacking → Isolation Forest anomaly detection → SIGMA rule enrichment → multi-agent LLM analysis. Zero-hallucination via hash-chained provenance. Kaggle AI Agents Capstone — Agents for Business.
Apache-2.0 -
Vetono update data
93 agentic MCP tools + 49 specialist agents for every major AI CLI — self-learning, no API keys. Veto gives Claude Code, Codex CLI, Gemini CLI, Cursor, Windsurf, Zed, and JetBrains a council of specialist agents, persistent cross-platform memory, a self-learning router, CI/CD gates, and dependency-hallucination + decision-drift guards — all powered by your existing AI subscription. No separate API keys, no extra billing. **Highlights** - **Council debate** — multi-agent GREEN/YELLOW/RED verdicts on risky changes (LLM-backed via MCP Sampling, with deterministic fallback) - **Cross-platform memory & session handoff** — save on Claude, resume on Gemini or Codex; nothing needs re-explaining - **Self-learning router** — improves task routing from recorded outcomes - **Guards & gates** — CI gate, diff review, secrets scan, dependency-hallucination verification, decision-drift enforcement - **Compact mode** — advertise 9 tools instead of 93 and save ~16K context tokens; every tool stays callable **Links**: [GitHub](https://github.com/jigyasudham/veto) · [npm](https://www.npmjs.com/package/@jigyasudham/veto)
smithery -
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 -
Behavioral safety, vulnerability probing, and product-risk auditing platform & MCP server for Generative AI & Large Language Models.
pypi Apache-2.0 ★ 2 -
MCP server that checks an AI agent's answer against its tool outputs before you see it: exact checks, NLI and an LLM judge, with block / annotate / rewrite policies
MIT -
mcp-hallucination-suitegithub.com/steveswain14/mcp-hallucination-suiteupdated this year
A unified suite of MCP suppressors that prevent hallucinations, enforce grounding, and stabilise agent behaviour.
NOASSERTION -
Rule-based guardrails for AI-written reports: Python fills the numbers, checks each sentence, and keeps replayable receipts. Library, CLI & MCP.
MIT -
eruka-mcpgithub.com/dirmacs/eruka-mcpupdated this year
MCP server for Eruka — anti-hallucination context memory for AI agents
cargo MIT ★ 3