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.
-
Model Context Protocol server with TypeScript for jailbreak detection tools
npm MIT -
Pattern-based prompt-injection and jailbreak screening with optional semantic analysis.
npm -
Autonomous Prompt Engineering, Jailbreak Defense, LLM Cost Optimizer & Eval Benchmark MCP for Cursor, Claude & Windsurf
-
A firewall for LLM traffic: detect prompt injection, jailbreaks, and secret/PII leaks. Zero-dependency Python library + CLI + MCP server.
MIT ★ 1 -
Give your AI agent a wallet it cannot drain: an on-chain allowance (Solidity), an MCP server, and a public jailbreak challenge on Base.
MIT -
rogue-live-redteamgithub.com/nguiasoren/rogueupdated this year
Continuous open-web LLM red-team: harvests live jailbreaks from 15+ open-web sources, reproduces them against your model x system-prompt x tools, and serves results over its own MCP server.
pypi MIT -
AI Safety HOT: daily AI safety news, papers and incidents in one place — jailbreaks, prompt injection, alignment, evals, governance. Website, daily digest, RSS and MCP.
NOASSERTION ★ 2,127 -
Guardrino by HisnLabs: the AI guardrail for Arabic, its dialects, Arabizi and English. Stops prompt injection and jailbreaks, masks Gulf personal data. OpenAI-compatible proxy, LiteLLM plugin, MCP security proxy. Built in Dubai.
Apache-2.0 ★ 1 -
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 -
MCP security scanner (CLI + MCP server): detects tool poisoning, prompt injection, and rug-pulls in MCP manifests. Also an OWASP LLM Top 10 prompt-injection/jailbreak CLI for LLM apps.
npm MIT ★ 3 -
Prompt-injection scanner for AI agents: deterministic, documented rules (instruction override, jailbreaks, exfiltration URLs, tool hijack/MCP tool poisoning, hidden Unicode, encoded payloads) returning a risk score, categories and spans. Library + x402 pa
npm MIT -
Mobile security toolkit for React Native: runtime root/jailbreak, debugger, hooking and integrity checks on Android and iOS, a static auditor with OWASP MASVS/CWE mappings and SARIF output, and AI-assisted review through your own model via MCP.
npm MIT -
AI-firewallgithub.com/Akhilucky/AI-firewallupdated this year
🛡️ AI Firewall — Multi-agent LLM security MCP server. Protects against prompt injection, jailbreaks, and policy violations.
pypi MIT -
Let AI agents operate a real iPhone. Local MCP server and agent, no jailbreak, no extra hardware.
AGPL-3.0 ★ 1,357 -
Run in-game button cheats on a stock PS5 with no jailbreak, over Remote Play. Web app, SQLite library and MCP server.
GPL-3.0 -
MCP server that lets an LLM agent see and drive a stock PS5 over Remote Play — no jailbreak, no capture card
pypi GPL-3.0 ★ 1 -
icemoon-mcpgithub.com/icemoon-dev/icemoon-mcpupdated this year
MCP server for controlling real, physical iPhones with AI — no jailbreak. Ships with the Icemoon app for macOS.
npm NOASSERTION -
Let Codex use your iPhone. Open apps, tap, type, and scroll over USB, with a live preview. No tracking or jailbreak.
MIT -
Give your Claude a thumb — open-source iPhone MCP. Tap, type, scroll, run skills through macOS iPhone Mirroring. No jailbreak.
MIT -
Sub-millisecond honeypot decoy MCP tools and tripwire defense against autonomous agent jailbreaks, unauthorized exfiltration, and prompt injections.
NOASSERTION -
Let any LLM drive a real iPhone — MCP server for the macOS iPhone Mirroring app plus Xcode/simulator/device test automation. 63 tools, Swift, no jailbreak.
mcpb MIT ★ 9 -
Let an AI model use a real iPhone with a $7 ESP32: AirPlay eyes, USB HID hands, any OpenAI-compatible vision model. No Mac, no developer account, no jailbreak.
npm MIT ★ 4,996 -
Let your AI agent use your iPhone. Open-source MCP server that lets Claude Code, Codex, Cursor, Gemini and any MCP agent see, tap, swipe and type on a real iPhone from your Mac. No jailbreak.
Apache-2.0 ★ 1 -
Prompt-injection defenses for Claude Code. A PreToolUse Bash hook blocks compositional credential-exfiltration shapes (secret read plus network, env dump to network, remote script to shell, reverse shells). A sanitizing MCP server wraps untrusted URLs and files in sentinels, strips invisible unicode, flags jailbreaks.
MIT ★ 4 -
让吃灰的 Kindle 重新活过来:MCP 推书、Apple Books 进度同步、高亮笔记汇总、Kindle 上读 AI Podcast(KOReader)
NOASSERTION ★ 3 -
A Local-First, Zero-Cost Prompt Injection Detection Server for the Model Context Protocol.
pypi MIT -
llamator-mcp-servergithub.com/RomiconEZ/llamator-mcp-serverupdated this year
MCP server for llamator: automate LLM red teaming workflows
CC-BY-SA-4.0 ★ 4 -
📲 One-tool MCP server for app work in a jailbroken vphone iOS VM: IPAs, tweaks, containers, regex-filtered logs.
GPL-3.0 -
PromptGuard Scan remote MCP documentation for prompt injection and LLM security testing.
NOASSERTION -
prompt-injection-mcpgithub.com/Xiangyu-Li97/prompt-injection-mcpupdated this year
MCP Server for authorized AI security research and penetration testing (Educational purposes only - requires explicit authorization)
NOASSERTION ★ 3 -
ps3-mcp-servergithub.com/metronoom82/ps3-mcp-serverupdated this year
MCP server for AI control of a jailbroken PS3 via webMAN MOD — 16 tools including PS3MAPI deep hardware access
MIT ★ 2 -
garbleworksgithub.com/SamsonCyber/garbleworksupdated this year
Authorized LLM red-team harness: recipes, scoped fire, evolutionary search, Wilson ASR. MCP + API + TUI.
Apache-2.0 -
⚡ C̷h̷u̷c̷k̷N̷o̷r̷r̷i̷s̷ MCP server: Helping LLMs break limits. Provides enhancement prompts inspired by elder-plinius' L1B3RT4S
npm ★ 69 -
mcp-prompt-suppressorgithub.com/steveswain14/mcp-prompt-suppressorupdated this year
A standalone MCP suppressor that filters, sanitises, and stabilises user prompts to prevent hallucinations and enforce grounding.
NOASSERTION -
Browser-based real-time modding tool for jailbroken PS5 consoles: find values, make cheats, debug, and learn how it is done.
GPL-3.0 ★ 5 -
promptarmor-plugingithub.com/allsmog/promptarmor-pluginupdated this year
LLM security red teaming — Claude Code plugin + MCP server with 80+ attack plugins and 25+ mutation strategies
MIT ★ 1 -
Prompt-injection guardrail with a hardened LLM judge, guarded RAG and an MCP server · FastAPI + React/TS
MIT -
让 Claude、Codex、Cursor 等 AI 助手看懂 iPhone 屏幕并直接操作手机的 iOS 自动化工具与 MCP Server,支持 UI 树获取、元素定位、截图与 OCR 识别、点击滑动、文字输入、应用管理和日志读取,适用于自动化测试与 iOS 逆向调试
MIT ★ 693 -
Connect AI coding tools to Seclai via Model Context Protocol (MCP) — manage agents, knowledge bases, and content sources from Claude, Cursor, Windsurf, and more
MIT ★ 1 -
pyrit-mcp-servergithub.com/Nislive/pyrit-mcp-serverupdated this year
Arm any MCP-capable agent to autonomously red-team a target LLM. Microsoft PyRIT exposed as 31 composable MCP tools: attacks, scoring, converters & scenarios
MIT ★ 1 -
Zero-dependency LLM red-team scanner: 24 attack probes across 9 classes (44/20 in PRO), honest refusal-aware verdicts, Markdown report, x402 USDC + Nano paid tiers
MIT -
shieldpi-mcpgithub.com/shieldpi1/shieldpi-watchtowerupdated this year
MCP server for ShieldPi Watchtower — run LLM security scans, generate drop-in guardrails from scan findings, operate the AI SOC analyst (incident triage + kill-chain reconstruction), and query the live agent monitor from Claude Desktop / Cursor / Continue.
pypi MIT -
Three-gate prompt rewriter for requests that keep getting refused: scrub input signatures, move intent to the instruction layer, shard the output past the filter. Ships a CLI, an MCP server, and installers for Codex / Claude Code / Claude Desktop / dsh.
npm MIT