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.
-
Test MCP Server with media handling and prompt capabilities for bootstrap testing
pypi MIT -
Testing AI-facing APIs: an own MCP server and an LLM endpoint on Claude with evals, groundedness and prompt-injection checks
MIT -
prompt-toolkit-mcpgithub.com/Money-Monkey-26/prompt-toolkit-mcpupdated this year
Production-tested AI prompt templates as an MCP server. 5 free prompts for code review, debugging, testing, and architecture.
MIT ★ 1 -
goodhabitz-evalab-mcpupdated this year
MCP servers split into 4 specialized services: Studio (content creation), Dev (prompt engineering), Test (QA automation), Admin (configuration)
npm ISC -
MCP server that gives Claude safe tools over a real web app, tested end to end through a real MCP client. Includes prompt-injection guardrails and 5 reusable Claude skills for QA.
NOASSERTION -
Agentic Work Assist - Production-ready MCP server with Constitutional AI Meta-Prompting Enhancement System featuring real-time prompt optimization, automated pattern recognition, constitutional AI-validated prompt evolution, A/B testing framework, compreh
npm MIT -
A lightweight MCP test server for verifying client connectivity, providing tools, resources, and prompts for integration.
-
Pulumi Prompts MCP Server (test)
npm MIT -
mcp-maestro-mobile-aigithub.com/krunal-mahera/mcp-maestro-mobile-aiupdated this year
MCP Server for AI-Assisted Mobile Automation using Maestro - Run mobile tests with natural language prompts. Features prompt analysis, tiered clarification, and enterprise-grade security.
npm MIT -
GenAI QA & Evaluation framework for RAG systems: API, Playwright UI, RAG evaluation, prompt-injection tests, a Python MCP server, and GitHub Actions CI
-
Community-driven library of tested prompts — npm for AI agent configs
pypi MIT -
sentinelmcpgithub.com/MusaIslamFahad/sentinelmcpupdated this year
An automated red-teaming and reliability-auditing platform for AI agents - tests for prompt injection, tool hijacking and data exfiltration. Ships as an MCP server.
MIT ★ 2 -
playwright-wizard-mcpgithub.com/oguzc/playwright-wizard-mcpupdated this year
MCP server providing Playwright test generation wizard with intelligent prompts and best practices
npm MIT ★ 7 -
mcp-prompt-labgithub.com/nikolasbarwicki/mcp-prompt-labupdated this year
Prompt evaluation workbench as an MCP server — run evals, compare providers, and test assertions directly from Claude Code or any MCP host.
MIT -
PromptGuard Scan remote MCP documentation for prompt injection and LLM security testing.
NOASSERTION -
Lets AI coding agents write and run Appium mobile tests from Markdown prompts, using Appium MCP
MIT -
@cqsjjb/meter-sphere-mcp-serverhttp://192.168.1.242:10985/jjb/ai-agent/mcp/meter-sphere-mcp-server.gitupdated this year
MCP server for MeterSphere test cases platform - Get test cases, generate AI test prompts, track testing progress, and execute browser automation tests
npm ISC -
AI-driven Android & iOS app testing over MCP: explore from a prompt, verify checkpoints, keep the evidence.
-
Enterprise-grade dynamic prompt management system with runtime prompt application, enhanced MCP server, constitutional AI, workflow orchestration, session management, metadata handling, and comprehensive testing framework
npm MIT -
Testers.ai CARBON Test Harness Litegithub.com/jarbon/carbon-lite/blob/main/mcp/README.mdno update data
Local, MIT-licensed AI test harness with 10 MCP tools and 11 prompts for risk-based test planning, focused testing, accessibility checks, evidence capture, screenshot maps and interactive HTML quality reports. Runs over stdio in one explicitly selected project. Your coding agent supplies AI and browser execution; CARBON records observed evidence and generates reports. No hosted backend, model API key or telemetry required.
smithery -
A lazy prompt fixes itself on open-weight models (Gemma 3, gpt-oss) over AcruxCore's MCP server. Banking77, held-out exact-match test.
Apache-2.0 -
Security auditor for AI agents and MCP servers — config analysis, prompt injection testing, tool probing, data flow tracing
npm MIT -
The developer toolkit for testing, analyzing, optimizing, and versioning AI prompts. (CLI, LSP, MCP Server, VS Code)
MIT ★ 1 -
combot-mcpgithub.com/GabrielBrasileiro/combot-mcpupdated this year
AI-powered MCP server that gives your assistant built-in knowledge of Combot, with no need to explain the library in every prompt to create your Android Compose tests.
★ 2 -
GuardXgithub.com/ai-craftsman404/GuardXupdated this year
LLM security scanner for Claude Code, with MCP-based testing for prompt injection, RAG risks, and supply-chain exposure.
MIT -
MCP server for QA automation: 10 tools, 4 resources, 3 prompts for insurance testing. Full Claude Code ecosystem.
MIT -
A guarded MCP server that lets Claude query a MySQL database safely: single SELECT only, allowlisted columns, row limits, prompt-injection tests and measured results.
-
Check a prompt-template pack from supplied templates: variables, versions, test renders.
npm MIT -
MCP server for generating automated shinytest2 test cases from natural language prompts
MIT ★ 1 -
Forward-deployed engineering in public: a role-scoped MCP server with prompt-injection defenses, 173 tests, and an eval harness (45.9% to 94.6% safe decisions). Live dashboard + threat model + ADRs.
-
Search 1,000+ battle-tested AI prompts from getprompts.org into Claude Code, Cursor, ChatGPT.
npm -
devglidegithub.com/DannyNsITServices/devglideupdated this year
Dev tools your AI assistant is missing. Kanban, shell, tests, workflows, prompts, docs — all as MCP servers on one port.
npm MIT ★ 2 -
Agent-operable expense-approval console built on WebMCP. A ChatGPT browser agent triages an approval queue under server-enforced policy, with role-scoped tools, server-authorized money actions, and prompt-injection containment proven by tests.
MIT -
RAG assistant for SEC filings that cites its pages. Evaluated on 150 FinanceBench questions and stress-tested on amended filings, wrong-company pages and prompt injection, across three models. LangChain, LangGraph, Chroma, FastAPI, MCP, Google ADK.
MIT -
playvlay-mcpactive
Generate playable, play-tested HTML5 games from a prompt. MCP server for PlayVlay.
npm MIT -
The crash test for AI coding agents: 19 realistic traps for Claude Code, Codex and Gemini CLI in a sealed container, plus a safety grade with fix prompts for apps built with Lovable, Bolt, v0 and Replit.
npm Apache-2.0 ★ 1 -
AI-Act-Evidence-Deskgithub.com/MoSaleh47/AI-Act-Evidence-Deskupdated this year
Agentic RAG assistant that turns an AI use case into a source-grounded EU AI Act evidence map. Hybrid BM25/TF-IDF retrieval with rank fusion, three-pass self-consistency synthesis, MCP server, RAGAS evaluation and prompt-injection tests.
-
mcp-quickstartgithub.com/Rustamovppl/mcp-quickstartupdated this year
A small, tested finance-toolkit MCP server in Python (tools + resource + prompt, with CI)
-
bestaiimageprompt-mcpgithub.com/rocnubie/bestaiimageprompt-mcpupdated this year
Best AI Image Prompt (https://bestaiimageprompt.com) is a curated gallery of tested AI image prompts, each paired with its actual rendered output. Every entry is sourced from working creators on X, Reddit, and Discord — real strings that real people pasted into real models and published.
MIT -
A FastAPI and MCP server connected to Supabase table for managing, testing, and programmatically accessing LLM System Instructions, System Prompts, BOTs, Google Gems.
★ 14 -
promptlintgithub.com/cognis-digital/promptlintupdated this year
Lint, version, and test prompts as code with a CI gate
NOASSERTION -
proper-perplexity-mcpgithub.com/minanagehsalalma/proper-perplexity-mcpupdated this year
Hardened Perplexity MCP fork with image uploads, structured outputs, MCP prompts/resources, and real stdio test coverage.
MIT -
agentic-dev-assistantgithub.com/SaifulJnU/agentic-dev-assistantupdated this year
Multi-agent AI assistant that turns a GitHub issue into a draft PR with tests; planner/coder/reviewer loop on LangGraph, served via FastAPI + MCP. Swappable Claude / GPT / local Hugging Face models. Prompt caching & model tiering for ~50% lower cost. Dockerized, CI-tested.
-
rule-based-verifiergithub.com/kroq86/rule-based-verifierupdated this year
Human-led workflow for AI-assisted coding: policy first, small steps, verification (tests, lint, search) via MCP—not bigger prompts.
-
dox: tests how AI agents behave under adversarial inputs (duplicate refunds, prompt injection via tool output). Deterministic PASS/FAIL evaluator. Built on TrueForge + MCP.
-
Tool-calling SQL agent with read-only guardrails, an MCP server and an evaluation harness (bootstrap CIs, ablations, prompt-injection test)
-
payments-agentgithub.com/alex-jb/payments-agentupdated this year
Overdue-invoice reminder drafter. Stripe-shaped types + MockProvider fallback. Hard money-safety prompt rules (verified in tests). pip · MIT.
pypi MIT ★ 2 -
agileplus-cligithub.com/phenotype/agileplusupdated this year
Typed intent graphs for engineering work — convert prompts into validated DAGs of intents, features, tasks, tests, and edges, then store, query, and serve them from Rust, HTTP, or MCP.
cargo MIT -
Multimodal RAG agent over CC-BY clinical-trial papers: MCP retrieval tools, A2A reviewer agent, OpenTelemetry tracing, chunking evaluation and prompt regression tests
MIT -
mcp-document-chatgithub.com/JuanCamiloMendoza99/mcp-document-chatupdated this year
A Claude CLI chat client backed by a Model Context Protocol server — FastMCP tools, resources and prompts, with a 27-test suite.
-
arthagithub.com/sidhvik17/arthaupdated this year
Multi-agent equity research crew (Google ADK) that writes balanced briefs and refuses financial advice by construction - deterministic prompt-injection guardrail, MCP tool server, 12 tests. Kaggle x Google Agents Capstone.
MIT ★ 2 -
mcp-memugithub.com/devbysatyam/mcp-memuupdated this year
🤖 MCP server for AI-powered MEmu Android Emulator control — 69 tools, 7 prompt templates, 8 resources for VM lifecycle, app testing, UI automation, GPS spoofing, and more
MIT ★ 2 -
standard-vocal-mcpupdated this year
Voice Agent Factory MCP — deploy, eval, and audit phone agents built on Vapi. Differs from the official Vapi MCP: vertical templates, self-testing agents, audio forensics, prompt versioning, CI regression gates.
npm MIT -
An AI trading agent that is never allowed to make a trading decision. Deterministic Python core. Every rule is in tested code, not prompt text.
MIT ★ 2 -
MCP server and agent skill for SEO and AI search visibility: audit AI crawler access, generate llms.txt, and build GEO/AEO prompt tests.
npm MIT ★ 2 -
Save points for your project before each Claude Code prompt and before destructive commands like rm -rf. Finds the newest one your tests pass and recovers it to a new branch. Local, MIT.
MIT -
Multi-tenant "ask your documents" app: FastAPI, pgvector hybrid search, OCR, background jobs, a LangGraph agent with an MCP tool server, role-based access, prompt-injection defenses, and CI. Measured, tested, and deployed.
-
MCP server built with FastMCP and the Kaggle API that lets Claude search, inspect and download Kaggle datasets using plain-English prompts. Includes three tools, MCP Inspector testing and Claude Desktop setup.
MIT -
LangGraph agent that reads leases, insurance policies, and ToS and traces every conclusion to its source clause. Custom MCP retriever; prompt versions and chunking strategies benchmarked on a hand-labeled test set. Live Streamlit demo.
MIT -
Stealth browser automation for AI agents — unattended, headless, N parallel persistent logged-in Chromes. One MCP server + CLI, credential vault with TOTP/IMAP 2FA, vision clicking, prompt-injection scanning. Self-hosted, real Chrome. 2,027 tests.
pypi Apache-2.0 ★ 9