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.
-
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 -
HALO (Harness for Always-On Loop Oversight), from the survey Always-On Personal Intelligence: an evaluation harness and MCP plugin for always-on personal AI assistants and LLM agents. It checks whether Sensing, Memory, and Action are connected and whether self-evolving agents learn from outcomes, via paper and code audits and ablations.
MIT ★ 8 -
The experience graph for AI agents. Self-hosted memory, human-reviewed decisions, and outcome-backed datasets for model evaluation and learning.
Apache-2.0 ★ 23 -
Give your AI agents an evaluation tool - provider-agnostic LLM eval runner with an MCP server for Claude, structured output validation, datasets, evaluators and honest cost metrics
MIT -
Local-first MCP server for versioning, searching, evaluating, and improving AI prompts with coding agents.
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 -
An integrated AI platform composed from nine of these libraries: an authenticated multi-tenant API, OpenAI and Anthropic failover, streamed agent steps, an MCP server, per-tenant memory, durable multi-agent jobs with crash recovery, tracing, and evaluation.
MIT -
MCP Server implementation of AgentFold - Long-Horizon Web Agents with Proactive Context Management. v2.0 adds analytics, validation, quality evaluation, and performance benchmarking.
npm MIT -
Universal Model Context Protocol (MCP) server and CLI runner for TypeSafe Jev decision models. Calibrated evaluation for code reviews, acceptance gating, and agent guardrails across Claude Code, OMP, and Cursor.
MIT ★ 2 -
governed-contextgithub.com/leonardoeverling-spec/governed-contextupdated this year
Governed memory for AI agents: explicit states, fail-closed sensitivity gates, quarantine-only writes (dependency-free MCP server) + a blind-evaluation protocol skill
MIT -
ML Referee evaluates whether candidate features add predictive value beyond a baseline for machine learning and forecasting. AI agents can preflight data, request quotes, and run authorized evaluations through MCP. Invite-only beta. Website: MCP endpoint: https://api.mlreferee.com/mcp
MIT -
Omneum: Agent Context Reliabilityno update data
Evaluate source reliability, dependent evidence, and conflicting claims in your agent's context.
smithery -
agent-negotiation-mcpgithub.com/csoai-org/agent-negotiation-mcpupdated this year
AI-powered agent negotiation MCP server for agents. Supports propose deal, evaluate offer, counter offer. By MEOK AI Labs.
pypi MIT License -
MCP server for Nia AI WebEvalAgent - Web automation and evaluation agent
npm MIT -
NIA AI Web Evaluation Agent MCP Server - Autonomous browser testing and debugging
npm MIT -
A high-fidelity, local-first MCP server and protocol for storing, versioning, syncing, and evaluating AI agent prompts and skills
MIT ★ 3 -
MCP Server for Reftrix - AI agent integration for web design analysis, layout extraction, motion detection, and quality evaluation
npm AGPL-3.0-only -
An evaluation infrastructure layer for autonomous coding agents. Measure true FinOps costs, cache economics, and risk-aware routing for Claude Code.
-
SLAC-MCPgithub.com/DennisPrediger/SLAC-MCPupdated this year
MCP Server to expose evaluation Tools to LLM and AI Agents using a Simple Logic & Arithmetic Compiler
Apache-2.0 -
hive-market-mcpgithub.com/atomicintuition/hive_marketplaceupdated this year
Hive Market MCP Server — discover, evaluate, and install MCP tools for AI agents
npm MIT -
path402github.com/b0ase/path402updated this year
The $402 Protocol — MCP server for AI agents to discover, evaluate, and acquire tokenized content
npm MIT -
@agentdance/node-webrtc-mcpgithub.com/agent-dance/node-webrtcupdated this year
MCP server for @agentdance/node-webrtc — lets AI agents discover, evaluate, and get started with the pure-TypeScript WebRTC stack
npm MIT -
Real-time tool usage tracking, agent telemetry, progress tracking, and automated commit-readiness evaluation MCP server.
-
Usable Browser Agent, free personal/evaluation tier: an MCP server plus browser extension that lets your AI agent drive your real, logged-in browser.
npm SEE LICENSE IN LICENSE.txt -
Exact arithmetic for AI agents. Two independently written engines (Python exact rational/decimal + a separate C++ evaluator) cross-check every result.
MIT ★ 2 -
All-in-one open-source toolkit for reliable AI agents: memory, code search, safe execution, and evaluation in one npx install.
MIT ★ 3 -
mcp-ai-agent-guidelinesgithub.com/anselmoo/mcp-ai-agent-guidelinesupdated this year
MCP server exposing public instruction workflows as tools, backed by hidden AI agent skills for requirements, orchestration, quality, research, evaluation, governance, resilience, and physics-inspired analysis
npm MIT -
saigegithub.com/urmzd/saigeactive
Go SDK, CLI and MCP server for streaming agents, tools, memory, retrieval and evaluations, with pluggable model providers and durable execution.
Apache-2.0 ★ 6 -
recourse-cligithub.com/recourseos/recourseupdated this year
MCP server for AI agents to evaluate consequences before destructive actions. Analyzes Terraform plans, shell commands, and MCP tool calls.
npm MIT -
dollar402-mcp-servergithub.com/b0ase/dollar402-mcp-serverupdated this year
MCP server for the $PATH protocol — AI agent tools for discovering, evaluating, and acquiring tokenised content via $402
npm MIT -
🧭 Strategic BDI cognitive reasoning engine for AI agents. Goal decomposition DAGs, multi-factor utility scoring, risk evaluation, and adaptive replanning over MCP.
npm MIT ★ 13 -
An MCP server exposing passmcp's diagnostics as read-only tools: evaluate an MCP server, or verify an attestation, from inside the agent.
GPL-3.0 -
MCP server that sits between AI agents and real systems. Evaluates every tool call before execution. Holds risky ones for human review.
Apache-2.0 -
Guarded AI operations agent using OpenAI Codex, MCP, FastAPI and SQLite with human-approved write actions, audit logging, webhooks, and automated evaluations.
-
MCP server giving AI agents full debugger control over running Java applications — inspect state, set breakpoints, evaluate expressions, and mutate values at runtime via JDWP/JDI
MIT ★ 34 -
Strategic BDI reasoning engine for autonomous AI agents — goal decomposition, utility scoring, risk evaluation, and reactive replanning over PuterVision memory and world models.
npm MIT -
SewaSetu AI - Autonomous Agentic Public Welfare & Pension Portal by Harish featuring dual MCP servers, RFC 9728 discovery, OAuth 2.1 privilege segregation, and live persona evaluation framework.
MIT -
MCP server for autonomous AI agent trust — cryptographic provenance (Chain of Consciousness with Bitcoin OpenTimestamps), decentralized reputation (Agent Rating Protocol with bilateral blind evaluation). Part of the 7-protocol Agent Trust Stack.
pypi Apache-2.0 -
Reasoning orchestration MCP - Tree of Thoughts with multi-agent evaluation. Fork of Anthropic's sequential-thinking with string coercion fix.
npm MIT -
@iflow-mcp/ssdeanx-deep-researchgithub.com/ssdeanx/deep-researchupdated this year
Advanced AI deep research assistant using Mastra's workflows and agent capabilities. Creates an interactive, human-in-the-loop research system that allows users to explore topics, evaluate results, and generate comprehensive reports.
npm ISC -
mcpgithub.com/decionis/mcpactive
Decionis MCP policy tools and native Claude Code, Codex, and Copilot execution hooks — evaluate agent actions against DECIONIS_POLICY.md before they run. npx -y --package=@decionis/mcp decionis-mcp
Apache-2.0 -
Cortex-aigithub.com/CharonK652302/Cortex-aiupdated this year
Production RAG app — LangChain LCEL, LlamaIndex, Crew.ai multi-agent, FAISS, FastAPI, Docker, MLflow evaluation, GitHub Actions CI, MCP server — zero paid APIs
-
An agentic AI–powered IaC scanner that evaluates cloud infrastructure against security and compliance benchmarks. It also delivers cost analysis and AWS Well-Architected recommendations, with support for custom policies, checks, and configurable exceptions.
Apache-2.0 ★ 4 -
mindcore-memory-mcpgithub.com/woshilaohei/mindcore-memory-mcpupdated this year
Production-hardened MCP memory server for AI agents: hybrid BM25+FAISS search, circuit breaker, SLO tracking, Prometheus metrics, BND boundary evaluation. Gives AI persistent long-term memory that survives across sessions.
MIT ★ 1 -
MCP server for Agent Referee: free demo tie-breaks between AI agents, sealed review of an agent's own output (free daily allowance, then paid), and offline checks of signed receipts. For technical evaluation of AI agents' work; results are advisory.
npm MIT -
Per-agent-pair IAM for A2A. Define policies ('orchestrator may call billing only when amount<1000'), gate every A2A call via evaluate_call. EU AI Act Art 14 + ISO 42001 Annex A.7 evidence with signed policy-decision attestations.
pypi MIT License -
Unit converter for AI agents, powered by the GNU units database — convert 3000+ units of measurement, evaluate compound unit expressions, reduce to SI base units, dimensional analysis. Offline, deterministic.
GPL-3.0 ★ 1 -
Anistroph - multi-domain predictive analytics architecture for temporal and non-temporal modeling, explainability, multidimensional evaluation, and MCP-based inference with Claude and AI agents.
GPL-3.0 -
Stadium & arena ticketing built as a reference-grade Clean Architecture solution - .NET 10, DDD, CQRS, transactional outbox/inbox, optimistic concurrency, Keycloak, Elasticsearch, Stripe, .NET Aspire - with an AI layer: an MCP server, an evaluated agent and a RAG policy assistant on pgvector.
MIT -
md-feedbackgithub.com/yeominux/md-feedbackupdated this year
MCP server for markdown plan review — companion to the MD Feedback VS Code extension. AI agents read annotations, mark tasks done, evaluate quality gates, and generate session handoffs. 27 tools for Claude Code, Cursor, and other MCP-compatible clients.
npm SUL-1.0 -
tuningengines-cligithub.com/cerebrixos-org/tuning-engines-cliupdated this year
Tuning Engines CLI, MCP server, and Python agent runtime adapters for governed model, agent, skill, and MCP workflows. Fine-tune open-source LLMs, run inference, manage datasets/evaluations, and connect LangGraph or Temporal while Tuning Engines handles p
npm MIT -
mcp-rag-agentgithub.com/luisrodriguesphd/mcp-rag-agentupdated this year
This repository presents a production-ready RAG system combining LangGraph agent with Model Context Protocol (MCP) integration. Features hybrid search using Reciprocal Rank Fusion (RRF) via MongoDB vector and text searches, grounded responses using COSTAR prompting, and automated RAGAS-based evaluation for building reliable, context-aware AI agents
★ 4 -
AI-Credit-Underwriting-Engine-using-FastMCP-LangGraphgithub.com/SayamAlt/AI-Credit-Underwriting-Engine-using-FastMCP-LangGraphupdated this year
Successfully developed a state-of-the-art AI-driven credit underwriting engine using LangGraph and FastMCP that orchestrates multiple specialized agents to evaluate risk, detect fraud, and generate explainable credit decisions, delivering transparent, auditable, and tailored financial solutions with unmatched efficiency.
Apache-2.0 ★ 6 -
Machship MCPno update data
# Machship MCP Server Connect your AI assistant to the Machship freight platform and manage shipping operations directly through natural language. Machship MCP enables AI agents to interact with freight workflows including consignments, quotes, carriers, labels, manifests, locations, and business entities. ## What you can do - 📦 **Consignments** — create, update, search, track, and manage shipments throughout their lifecycle - 💰 **Quotes** — retrieve carrier pricing, compare services, and evaluate transit options - 🏷️ **Labels & Manifests** — generate shipping labels, create manifests, and manage carrier bookings - 🚚 **Carriers & Services** — discover available carriers, routes, and delivery options - 📍 **Locations** — search, validate, and manage pickup and delivery addresses - 🏢 **Companies & Identities** — manage organisations, users, and account relationships - 📎 **Additional workflows** — handle attachments, invoices, custom schemas, and shipment consolidation ## Coverage **125 tools across 21 modules** providing broad access to the Machship API through a single MCP server. ## Compatible clients Works with any MCP-compatible client, including: Claude Desktop · Cursor · VS Code · Windsurf · Goose · Warp · LM Studio ## About Built by [Devkind](https://devkind.com.au) — delivering modern software solutions and AI-powered integrations for businesses worldwide.
smithery -
the-metis-fca-handbook-ai-harness-mcpno update data
# Metis FCA Handbook AI Harness — MCP Server An MCP (Model Context Protocol) server that integrates the Metis FCA Handbook AI Harness into Claude Desktop and other MCP-compatible agents. **Tool name:** `evaluate_fca_handbook_applicability` Use this tool to analyze FCA Handbook applicability for a product, service, or firm. Returns structured compliance reasoning with verbatim handbook citations. ## Why choose The Metis FCA Handbook AI Harness? - **Efficiency** Multiplies effectiveness of compliance advice. Saves £80+ in token fees per Harness run. - **Deals with "Hard Problem"**, in which LLMs' token incentives prioritise training data so results are unreliable. - **Verbatim citations** Quotes verbatim entries from the FCA Handbook. Other AI systems struggle to do this. - **Matches real-world need** You do not have to start off knowing which sections you are looking for. Carries out structured searches across all 10,000+ FCA Handbook entries. - **Secure Design** Harness compartmentalisation, one-shot structure, and statelessness fits natively with emerging AI agent security standards such as OWASP Top 10 for Agentic Applications 2026. - **AI Accessible** Integrates easily with AI workflows and agents. Your LLM can use it as a tool. ## Design Principles The Harness is built on proven principles: - **Stateless** — Each request is independent; no session coupling - **One-shot** — Complete analysis in a single call; no multi-turn state - **Clear contract** — Explicit input/output schemas for easy integration - **Hosted** — Single canonical source; no version drift or stale data ## Use Cases ### Compliance AI Platforms Embed FCA reasoning as a service within your compliance platform. Users ask natural-language questions; your platform calls the Harness and presents structured reasoning. ### AI Agent Workflows Agents building compliance workflows can include FCA Handbook reasoning as a composed tool—no external API calls, just MCP configuration. ### Compliance Review Automation Integrate into document review or due-diligence pipelines. Automatically screen new rules against FCA applicability. ## For Regulatory Screening FS firms screening MCP servers will find: - **Simple model:** Stateless, no hidden state, no background jobs - **Clear contract:** Explicit input schema, output schema, error modes - **Transparent updates:** Version pinning; no automatic upgrades - **Single point of failure:** If the Harness is down, it is obvious; no cascading config issues - **Audit-friendly:** All calls logged centrally, not distributed
smithery -
GenPark Reasoning MCTS Step Verity Evaluator - Implements Monte Carlo Tree Search (MCTS) Step-Level Process Reward Modeling (PRM) & Hallucination Branch Pruning.
MIT -
Stop shipping agents on vibes. Score every agent output for quality, safety, and cost.
npm MIT ★ 10 -
Quorel is an MCP server that scores AI agent responses before they reach a user, a trust layer an agent calls mid-turn, not a dashboard someone checks after the fact.
NOASSERTION -
MO§ES™ — Enterprise AI Operator Evaluations. Marketing site + MCP server. mos2es.org
NOASSERTION -
美股投研 Agent 原型:通过 MCP 提供行情、SEC 财报检索(RAG 带引用)、财务数据、组合风险和模拟下单工具,下单必须人工确认;内置 Agent 评测(任务成功率、工具调用准确率、成本与延迟)。
MIT ★ 1