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.
-
adversarial-geopolitical-equilibrium-mcpgithub.com/apifyforge/adversarial-geopolitical-equilibrium-mcpupdated this year
Adversarial geopolitical equilibrium analysis for AI agents via the Model Context Protocol.
MIT ★ 1 -
Remote MCP server for bounded adversarial checks during AI assistant workflows, with caller-side evaluation.
MIT -
⚔️ Stress-test your product and startup ideas before writing code. An adversarial AI tool for multi-role debate, synthetic user feedback, and validation planning.
npm MIT ★ 4 -
🔌 Exposes MITRE ATLAS (adversarial AI) threat landscape over MCP — MIT
pypi MIT -
Multi-model adversarial plan review — orchestrates AI agents to critique and refine implementation plans
npm MIT -
tenth-man-mcpgithub.com/keeganthewhi/tenth-man-mcpupdated this year
Adversarial review system for autonomous AI coding agents. 3 contrarian agents challenge every critical change before execution.
-
Adversarial Co-Generation Engine — concurrent Builder+Breaker agents for production-grade AI code security
pypi MIT ★ 1 -
AI Guardian that challenges actions before execution. Forces clarity of thought through adversarial questioning and NLI-based reasoning evaluation.
-
critic-orchestratorgithub.com/aureliocpr-ctrl/critic-orchestratorupdated this year
MCP server that spawns adversarial reviewers in parallel (Claude CLI, or any OpenAI-compatible / Anthropic API) to verify a claim/fix. Anti-confabulation layer for AI coding agents.
pypi MIT -
debate-mcpgithub.com/grzesir/debate-mcpupdated this year
Stress-test your decisions with adversarial AI debate. MCP server for Claude Code. Skeptic vs Steelman, web-search grounded, confirmation bias attack.
npm MIT ★ 3 -
Disciplined Engineering Harness for AI Agents. MCP server + 21 adversarial skill gates. Real-time ground truth for React, Next.js, shadcn/ui, Motion, React Flow, Echo, Go, Rust, and Design Tokens.
★ 1 -
MCP server for Sliver C2 - Adversary emulation framework integration for AI assistants
MIT ★ 9 -
Pre-execution decision verification for AI agents. Sentinel → adversarial RV panel (SERV Reasoning) — ALLOW / BLOCK / UNCERTAIN with signed attestation. MCP server, x402-gated on Base.
npm MIT ★ 1 -
cardcataloggithub.com/ddnet-repo/cardcatalogupdated this year
The certification authority for AI agents. Adversarial exams, Ed25519-signed credentials, examiner economy. https://cardcatalog.ai
MIT -
Apexgithub.com/DikshithPulakanti/Apexupdated this year
Autonomous multi-agent AI research scientist that discovers hypotheses, debates them adversarially, and drafts patent claims — orchestrated by 4 custom MCP servers over a Neo4j knowledge graph.
-
govbenchgithub.com/CSOAI-ORG/govbenchupdated this year
🔌 Lets you measure any AI's refusal behaviour on an adversarial grid — reproducibly, for free.
Apache-2.0 -
Policy checkpoint between AI agents and PayPal refunds: MCP security proxy, LangGraph risk pipeline with human-in-the-loop approval, adversarial evals.
MIT ★ 1 -
dox: tests how AI agents behave under adversarial inputs (duplicate refunds, prompt injection via tool output). Deterministic PASS/FAIL evaluator. Built on TrueForge + MCP.
-
MCP server for Claude Code — a quorum of AI models (Codex, Gemini, Claude Fable with fresh context) for standard + adversarial review and consultation, synthesized by Claude Code
npm MIT -
Standalone Model Context Protocol (MCP) Server providing real-time cognitive research registry queries, EU AI Act compliance auditing, and adversarial human-AI interaction benchmarks.
npm MIT -
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.
-
Open-source AI security platform providing perimeter defense for LLMs and AI agents through swarm analysis, policy enforcement, adversarial testing, and real-time threat detection.
pypi AGPL-3.0 ★ 21 -
evidencegene-courtgithub.com/FUYOH666/evidencegene-courtupdated this year
Adversarial autonomous DFIR: Prosecutor/Defender/Arbiter AI court over a typed read-only MCP server. Findings without real evidence refs are structurally rejected. SANS FIND EVIL! 2026.
MIT ★ 1 -
mcp-guardrailgithub.com/mini4ai4/mcp-guardrailupdated this year
An MCP server that gives an AI agent read access to a database without giving it the database — enforced by SQLite's authorizer, proven by an adversarial fuzz harness.
MIT -
Task-context compiler for AI coding agents — TypeScript AST retrieval, bounded source slices and safe symbol patches over MCP. Content-hash preconditions, atomic multi-file patches with rollback, and an adversarial security suite. 111 tests, CI on Linux/macOS/Windows.
MIT ★ 1 -
Language-agnostic Spec-to-Contract AI Testing Harness & 500 Crash Preventer. Fuzzes OpenAPI/JSON Schema endpoints with boundary & adversarial vectors, provides zero-token deterministic replays, and integrates seamlessly via MCP for AI Agent ecosystems.
Apache-2.0 -
claude-mcp-bridgegithub.com/JaimeJunr/claude-mcp-bridgeupdated this year
Let any AI agent (Cursor, Codex, Gemini/Antigravity) delegate to Claude Code via MCP: adversarial code review, validation & fully autonomous coding tasks — with model + effort control and session follow-up.
MIT ★ 2 -
secondopinion-x402updated this year
Adversarial verification for AI agents: pay an independent skeptic per verdict in USDC via x402 (Base, Algorand, Solana). Structured verdicts with signed, offline-verifiable receipts. Vercel AI SDK + LangChain adapters, npx MCP server, free trial — no wal
npm MIT -
clonstgithub.com/capritora/clonstupdated this year
Adversarial AI code review for Claude Code: an MCP server that has your plans and code critiqued by OpenAI Codex (via your ChatGPT subscription, no API keys), in a ping-pong loop until consensus.
MIT ★ 1 -
O-1A visa evidence mapping: source-cited petition drafting, an adversarial USCIS-officer agent grounded in real AAO decisions, attorney sign-off, and evals. Next.js 16, Supabase, Vercel AI SDK, Mastra, MCP; runs on local models at $0.
MIT -
Production-grade AI coding toolkit — 71 agents (incl. devil-mode adversarial crew), 196 skills, 104 commands, parallel multi-agent commands, semantic intent routing, self-learning memory, and a built-in MCP server (16 tools / 6 prompts / 389 resources) th
npm MIT -
telosgithub.com/nocoder26/telosupdated this year
The AI that asks why. A Claude Code plugin that refuses to let founders write product code until they've validated with real customers. 13 skills (concierge onboarding, adversarial customer roleplay, optional gbrain bridge) + brain MCP. Free, MIT.
MIT -
Autonomous cyber red team intelligence for AI agents — this MCP server gives Claude, GPT-4o, and any MCP-compatible agent the ability to run quantitative attack simulations, synthesize exploit chains, forecast vulnerability emergence, and model adversary behavior using eight rigorously implemented mathematical frameworks.
MIT ★ 1 -
agent-robustnessno update data
Run an AI agent against 5 adversarial attacks (credential leak, metric fabrication, false halt, runaway spend, log falsification), scored on what the harness observed it do '—' never its self-report. Free local self-diagnostic: runs on your own machine, in a temporary working directory, not a security sandbox. Optional paid tier runs the same five server-side in fresh single-use E2B microVMs (isolation enforced by E2B) and returns an ed25519-signed badge. Each attack is PASS, FAIL, or INDETERMINATE; an attack the agent did not demonstrably encounter is never a PASS, and a run containing an INDETERMINATE is never signed. A badge attests one artifact's observed behavior on one date. Not a safety certification.
smithery -
Executed-evidence verification gate for AI coding agents
pypi AGPL-3.0 ★ 4 -
adversarial-regulatory-arbitrage-mcpgithub.com/apifyforge/adversarial-regulatory-arbitrage-mcpupdated this year
Regulatory arbitrage detection and beneficial ownership tracing for AI agents via the Model Context Protocol.
MIT -
adversarial-corporate-opacity-mcpgithub.com/apifyforge/adversarial-corporate-opacity-mcpupdated this year
**Beneficial ownership detection and corporate opacity analysis** via the Model Context Protocol, built for AI agents that investigate entities across 6 international registries and 4 sanctions watchlists.
MIT -
BERGHAIN MCP server — sovereign AI stress-testing
★ 1 -
elenchus-mcpgithub.com/jhlee0409/elenchus-mcpupdated this year
Elenchus MCP Server - Adversarial verification system for code review
npm MIT ★ 5 -
MCP server that sends ideas to a local LM Studio model for adversarial review
MIT -
tenth-man-mcpupdated this year
10th Man Protocol — Adversarial review MCP server for Claude Code. Spawns 3 contrarian agents to challenge proposed code changes before execution.
npm MIT -
grok-mcpgithub.com/howardpen9/grok-mcpupdated this year
MCP server that wraps the xAI Grok CLI. Lets Claude Code, Cursor, Cline, and any MCP host use Grok as a peer code reviewer, adversary, and second-opinion consultant.
npm MIT ★ 11 -
agy-plugin-cc is a Claude Code companion for running Gemini CLI or Antigravity CLI (agy) as a cross-model task delegate and code reviewer, with pragmatic and adversarial review, MCP tools, and background jobs.
MIT ★ 14 -
Delegate tasks and reviews from any AI agent (Claude Code, Codex, Cursor, any MCP host) to any provider and model
Apache-2.0 -
arenagithub.com/tim101010101/arenaupdated this year
Multi-agent AI arena for debates, code reviews, and red-team challenges via Model Context Protocol (MCP)
npm MIT ★ 1 -
@ezark-publish/agentdesk-mcpgithub.com/rih0z/agentdesk-mcpupdated this year
MCP server for AgentDesk AI-to-AI Service Marketplace. Quality review, service catalog, and marketplace execution — all via MCP tools.
npm MIT -
Claude Code plugin for OpenAI Codex subagents, Codex Spark, parallel code review, and long-running Codex sessions via daemonless MCP
MIT ★ 7 -
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 -
blagithub.com/qybaihe/blaactive
🛡️ BLA (bla-agents / Blind LLM Agents) — Open-source Blue Defense MCP server & CLI protecting JS/TS, Electron ASAR, WASM, Python, Native Binaries, SQLite, GGUF/Safetensors & Android DEX against AI reverse-engineering agents (REA).
npm MIT ★ 5,904 -
world-model-mcp is a signed audit memory server for AI coding agents, delivered as an MCP tool. Every event Ed25519-signed and Merkle-chained; offline-verifiable against pinned public keys. Hosted platform with hybrid post-quantum signing at etch.systems.
go MIT ★ 24 -
claude-codex-teamgithub.com/ndcorder/claude-codex-teamupdated this year
MCP server that bridges Claude Code to GPT Codex for adversarial brainstorming, code review, and implementation planning
npm MIT ★ 2 -
Verifiable multi-agent voice-operations control plane with safety gates, MCP tools, a 2,048-case adversarial evaluation, and CI.
-
Capability-bounded Linux evidence: C++20 MCP authority boundary, bounded Python orchestration, and adversarial assurance.
Apache-2.0 -
tailtest-clinegithub.com/avansaber/tailtest-clineupdated this year
tailtest for Cline: MCP server + .clinerules pack + Memory Bank integration. Reaches 8+ editors via Cline. Adversarial test mode shipped from day one.
MIT ★ 1 -
Secure read-only MCP server for typed job search, deterministic ranking, explainable matching, adversarial defenses, and audited tool use.
MIT -
Give your coding agent an adversarial critic: an independent, fresh-session MCP reviewer for code and docs, with built-in convergence and any model you bring.
npm MIT ★ 2 -
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 -
Model Context Protocol (MCP) Server for O*NET 31.0 / ESCO v1.2.1 7-Primitive Occupation Resolution, 4-Gatekeeper ATS Verification, and 4-Seat Adversarial Hiring Committee Simulation.
MIT -
A2A v1.0 multi-vendor agent exchange: JWS-signed Agent Cards with pinned vendor keys, deterministic trust and quarantine, cost- and reliability-aware broker over A2A, MCP tools, metered local LLMs, adversarial benchmark reported as n/N.
Apache-2.0 -
Claude Code plugin: a channel-based short-form/long-form social content pipeline — storyboard with adversarial quality gates, video production, HITL-approved publishing, and autonomous growth loops. Built-in MCP server, 46 tools.
Apache-2.0 ★ 1