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 judge for code written by AI: plain Python, rules against hallucinations.
pypi MIT -
judgesgithub.com/KevinRabun/judgesupdated this year
MCP server with specialized judges to evaluate AI-generated code for security, cost, scalability, cloud readiness, and best practices.
npm MIT ★ 7 -
mcp-as-a-judgegithub.com/OtherVibes/mcp-as-a-judgeupdated this year
MCP as a Judge is a behavioral MCP that strengthens AI coding assistants by requiring explicit LLM evaluations
pypi MIT ★ 17 -
Open source AI evaluation framework — LLM-as-judge + assertion-based evals for any AI app. CLI + MCP server.
npm Apache-2.0 -
企业级智能客服 AI 系统参考实现:三路意图融合 · MCP 工具调用 · 三级记忆 · LangGraph 多 Agent 编排 · Monitor 闭环 · LLM-as-Judge 评测。零密钥可运行,含两引擎对照数据与诚实记录。
MIT ★ 3 -
MCP server + CLI that gives any AI agent a tournament-grade Yu-Gi-Oh! judge, deck builder and market analyst. Beta: under active development, check for updates.
npm AGPL-3.0-only -
eval-integritygithub.com/ipezygj/evalgateupdated this year
Dependency-free statistical checks for AI eval claims (multiple-comparisons, judge-bias, resolution, fragility) — as a CLI and an MCP server an agent calls before trusting a benchmark number.
pypi MIT -
BestTestergithub.com/nshportun/BestTesterupdated this year
Production-grade Playwright + TypeScript QA framework with AI-powered testing, LLM-as-Judge evaluation, MCP server, 7 CLI agents, security fuzzing, CI/CD pipelines, Jira sync, and Slack reporting — zero-config, plug-and-play.
Apache-2.0 ★ 11 -
Immigration case intelligence for attorneys. Case tracking, judge analytics, AI predictions.
npm -
aperture-mcpgithub.com/jaysinailabs/aperture-mcpupdated this year
A commitment tripwire for AI-edited decision docs — MCP server + Python library. A signal, not a judge.
pypi Apache-2.0 -
MCP server for Rastro catalog operations — 26 tools for schema, items, activities, snapshots, AI mapping, judging, and image editing
pypi MIT -
A science game for AI agents: tune, infer, or out-evolve a living world, judged on held-out seeds.
npm -
research-pilotgithub.com/Marwaguerfel/research-pilotupdated this year
Autonomous AI research agent: parallel planning, RAG memory, MCP server, LLM-as-judge eval, Streamlit UI
-
annuma-study-companiongithub.com/AlirezaFakari/annuma-study-companionupdated this year
AI active-recall study agent for AnNuMa — MCP server, read-only guardrail, LLM-as-judge
★ 1 -
code-review-crewgithub.com/siddlpskmi/code-review-crewupdated this year
Autonomous multi-agent AI code reviewer with custom MCP server, agentic RAG, and LLM-as-judge evals
-
@presentation-md/mcp-servergithub.com/isatimur/presentation-mdupdated this year
MCP server exposing all 13 presentation-md tools (render/export/audit/judge/scaffold/share/import/preview) to any MCP-compatible AI agent.
npm MIT -
One memory for every AI you use. A judge checks every write; you decide the unsure ones.
AGPL-3.0 -
braintrust-litegithub.com/hongjieren/braintrustupdated this year
Multi-model AI consultation MCP for Claude Code — runs Claude, Codex, and Gemini in parallel for Judge-style synthesis
npm MIT -
MCP server for Jev AI (jev-ai.pro): classification, scoring, yes/no checks, action assessment, batched decisions and saved judges.
MIT -
Signed AI measurement (13-axis, deterministic judge). Measurement, not certification; monitored containment; UNMEASURED is never wrong.
pypi Apache-2.0 -
@skillprompting/mcpactive
MCP server for Skillprompting — AI-judged prompt engineering contests with SOL prizes. Lets any MCP-compatible agent browse rounds, submit entries, and check results.
npm MIT -
frontkitgithub.com/lakshmi930/frontkitupdated this year
MCP server + eval gate for four small AI tools. Deterministic checks and an LLM judge fail the build when model behaviour drifts.
-
Twelve choices. Then ask your taste about anything. Aesthetic taste as a file any AI can load, with a judge and an MCP server.
MIT -
Graphwright MCP server: lets an AI assistant plan, run and judge a captured API flow (repeat runs, speed check) through the Graphwright command-line runner.
npm MIT -
Free pitch roaster: five judges score your startup pitch 1-10, one verdict, one change. Deterministic rubric, no AI bill, keyless JSON API + MCP.
-
evalmcpgithub.com/gashel01/evalmcpupdated this year
Evaluation for AI agents — judge-based scoring and native RAG metrics (faithfulness, relevancy, context precision/recall). Python lib · CLI · MCP server · FastAPI.
Apache-2.0 ★ 1 -
fahaligithub.com/future-legends-inc/fahali-integrationsupdated this year
Market intelligence AI agents can verify - signed receipts, judged track record (misses included), Brier-calibrated forecasts. Adapters for LangChain, CrewAI, LlamaIndex, AutoGen, smolagents. MCP: mcp.fahaliai.com
pypi MIT -
Local AI agent for Apache JMeter: turns test documents into requirements, runs the plan and judges every run with evidence (verdicts, baseline diff, Prometheus) - MCP built in.
Apache-2.0 ★ 2 -
X (Twitter) research for AI agents: read any post without login, search with every hit checked on X, judge value apart from reach.
MIT -
Is this true? Verify a claim, URL, file, or AI answer: up to 20 claims (configurable) judged against web sources, with confidence, the evidence span, and the reason. CLI + MCP server.
npm Apache-2.0 -
An MCP server that checks an AI agent's draft answer against its sources: figures, dates, citations and must/may wording, sentence by sentence; an optional LLM judge reads the prose.
pypi AGPL-3.0-only -
negative-memorygithub.com/aliomraniH/negative-memoryupdated this year
An MCP server that remembers mistakes before your AI agent makes them. PostgreSQL + pgvector stores structured anti-patterns, hybrid search surfaces them during planning, and an adversarial Judge critiques plans against known failures — shifting from reactive debugging to proactive immunization.
-
nocapgithub.com/xi029/nocapactive
No evidence, no answer. 🧢 System One evidence gate for RAG & AI agents: Jev / Laya / any LLM judges before your LLM speaks. MCP server, Python SDK, CI route tests.
Apache-2.0 ★ 1 -
Don't trust AI. Verify it. Open-source verification for evidence-grounded AI — deterministic grounding checks against source evidence, no LLM judge. CLI, GitHub Action, Python, MCP server, Docker.
go MIT ★ 1 -
Evidence-based YouTrack backlog prioritization for AI agents (Claude Code, Cursor): read-only MCP server + skills; the agent judges, a deterministic calculator computes, a human approves.
MIT -
Open-source, self-hosted observability for AI agents: OTLP ingestion, tracing SDK, sessions, cost analytics, datasets and LLM-as-judge experiments, alerts, replay, and an MCP server. SQLite or PostgreSQL.
pypi MIT -
MCP server for the Woodstitch API: paid human listeners judge your text-to-speech audio blind (pairwise preference or human-vs-AI detection). Not MOS.
pypi MIT -
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 -
Ticket tracker for AI and human work: Go + GraphQL + Postgres, SolidJS UI, and an MCP server so Claude Code agents can claim, file and work tickets, with LLM-as-judge evals.
MIT -
chance-mcpupdated this year
Chance verification harness as an MCP tool — the AI judge between an agent's intent and its action. Verify a proposed action against its mandate before it executes, and get a signed, independently-verifiable verdict.
npm MIT -
cdvgithub.com/azank1/cdvupdated this year
Conservative Dual-Verify — judging AI agent work with a deterministic floor + LLM critic (stricter score wins) and Bayesian stopping. Research preview, frozen at v1.0.1; active development continues privately. MIT.
pypi MIT ★ 4 -
Agentic-Document-Intelligence-Industrialized-Cross-Reference-Enginegithub.com/gayathrikumari/Agentic-Document-Intelligence-Industrialized-Cross-Reference-Engineupdated this year
🤖 Built a production-grade Agentic AI system using MCP + ReAct Multi-Agent Patterns to automate complex engineering document cross-referencing | Multimodal RAG | LLM-as-a-Judge | KV Caching
-
Ask questions about your PDFs and get answers with page-level citations. Hybrid FAISS + BM25 retrieval with Reciprocal Rank Fusion, LLM-judged evaluation (18/20 correctness, 20/20 faithfulness) and an MCP server for AI clients.
MIT ★ 1 -
Auditable, self-improving knowledge & memory for AI agents over MCP. Zero-key & multilingual by default · citation-enforced answers · replayable why-trace · judge-gated compounding loop. Postgres-only. Reference implementation of the Agentic Product Standard.
Apache-2.0 ★ 4 -
tardygradagithub.com/fabio-rovai/tardygradaupdated this year
A verified runtime for trustworthy AI agents: every claim passes an 8-layer verification pipeline before it counts. Deterministic checks, explainable verdicts, no LLM judging an LLM. Pure C, zero dependencies, MCP-native.
MIT ★ 26 -
AI shopping agents that can't spend what you didn't approve. Gemini agents negotiate, a deterministic policy engine judges every offer, your passkey signs the exact cart, and PayPal settles only after verification.
MIT -
realestate-ai-agentgithub.com/selvaonline/realestate-ai-agentupdated this year
DealSense — multi-agent AI platform for CRE private equity analysis. LangGraph.js supervisor + 6 specialists over 20 zero-token tools, domain-pack architecture, judge-grade evals, built-in MCP server. TypeScript/Angular/AWS.
MIT -
MCP server for driving Adobe Premiere Pro from an AI assistant, with tools that verify their own writes, measure colour and loudness, and judge whether an edit is ready to post
MIT ★ 1 -
listguard-mcpactive
Cross-border e-commerce listing compliance audit for AI agents. Deterministic rule engine (no LLM judging): banned words, absolute claims, certification risks, trademark risks and 2026 AI content disclosure. MCP server over stdio, zero dependencies.
npm MIT -
SentinelMind: the phone sentinel with a mind - AI-powered phone network patrol, the AI identifies, judges and explains your home network through your phone, audit-grade inspection, zero hardware. Python stdlib, Termux-ready.
MIT -
Kavach: a second listener that catches scam calls and warns elderly listeners in Telugu and Hindi. Built entirely on Gnani AI (Prisma, Evon, Timbre), with a code Judge, MCP family assistant and Bitcoin-sealed evidence.
-
Self-hosted AI web mail for all your mailboxes: an LLM judges what matters across inboxes and spam, pushes only that, and the web view opens on what needs you, to read, file and answer there. MCP server included.
MIT -
pulse-flight-deals-mcpgithub.com/yaotsakpo/pulse-flight-deals-mcpupdated this year
MCP server that lets an AI agent judge whether a flight price is actually a deal, grounded in observed fare history. Answers 'is $X for this route a good price?' with real numbers an LLM cannot produce on its own.
npm MIT -
Your stream of alpha: pulls from X, GitHub, Hugging Face, Substack, and papers, keeps only real signal with one hard judge, and turns it into project suggestions, tested agent skills, and an MCP server for your AI agents.
MIT -
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 -
AI asset-generation pipeline with a human in the loop: approve designs, AI builds editable Blender models, graphics, SVGs, PBR materials and sprites, judges the renders, and you accept. Self-hosted, with an OAuth-protected MCP server.
MIT -
plangithub.com/kylejtobin/planupdated this year
A control plane for trustworthy AI agents: write product judgment once, then enforce it through typed MCP tools, git gates, demolition-first stories, evidence-judged checkboxes, and zero-trust monitoring. Ships with Claude Code and Cursor plugins, Agent Skills, and a FastMCP server.
NOASSERTION ★ 5 -
Claudette: an AI that answers only from women's writing. Claude, like every LLM, was built mostly by men and trained mostly on text by men; this redresses the balance a little. Public-domain texts via Project Gutenberg, gated by Wikidata. Not perfect, judges nobody, meant to be interesting. MCP server, no key.
MIT -
Turn AI coding CLIs (Claude Code, Codex, Antigravity) into AI employees on LINE, Telegram, Discord, Slack and 7 more chat apps. A judge reviews each result, every cost is logged, memory stays on your machine. Also runs as a standalone MCP memory server. Self-hosted, Apache 2.0.
npm Apache-2.0 ★ 51 -
Repovive MCP Server connects AI assistants to competitive programming: browse problems and contests, solve and judge code locally, submit for real verdicts, track ratings, and author verified problems and rounds. Built with FastMCP v3 and httpx, with 93 tools, 10 resources, 8 prompts, Docker/Kubernetes deployment. Open source and extensible for AI.