- Conduid
- Marketplace
- #llm-security
MCP servers tagged llm-security
557 live MCP servers tagged "llm-security", ranked by trust score. Tags come from package metadata and repository topics, so this list covers servers across every category that work with llm-security.
Intercepter
Transparent proxy + enterprise gateway for MCP: audit logging, tool allow/deny policy, secrets/PII redaction, and OpenTelemetry for every AI agent <-> MCP tool call.
Agent Fence
Local security layer for AI coding agents that blocks secret leaks, destructive commands, unsafe Git operations, and unauthorized network access.
Memory Blackbox
Agent memory forensics — trace a bad agent action back to the exact poisoned memory, see the blast radius, roll it back
Prompt Foo
Promptfoo is an open-source CLI and TypeScript/Node.js library for evaluating, red-teaming, and security-testing LLM applications, agents, and RAG pipelines. It runs deterministic…
Arcjet Go
Runtime security for AI apps and agents: prompt injection detection, tool-call authorization, sensitive-data redaction, bot protection, and rate limiting. Drop it into your Go cod…
Threatveil Oss
Open-source assurance infrastructure for tracking whether security evidence still applies as autonomous AI systems change.
Agent Skill Guard
Security auditor for AI agent skills — catches hidden-Unicode prompt injection, reviewer-subversion, data exfiltration & post-install drift that markdown-only vetters miss.
Hecate Public
Security scanner and runtime guard for MCP servers and AI agents. Runs entirely on your machine.
Ccs Conformance Vectors
Public conformance test vectors for CCS (Correctover Conformance Shape) — signed, tamper-evident receipt fixtures for verifying AI agent runtime security implementations.
Agent Authority Check
FOXIFY deterministic authority preflight for MCP servers and AI agents: bind authority, inspect consequence, preserve evidence.
Redact Secret Adapters
Secret redaction integrations for Pino, Python logging, OpenTelemetry traces, MCP, and AI context. Powered by Redact Secret.
Semantic Intent Scanner
Semantic intent evaluation for agentic systems. Detects instruction-layer risks including authority laundering, consent bypass, concealed execution, and semantic attack patterns i…
LLM Redteam
面向中文场景的开源 LLM 红队测试工具 | An open-source red team toolkit for LLM applications, covering prompt injection, jailbreak and info leakage attacks, with OpenAI/Dify/MCP adapters and LLM-a…
Omnodex
Omnodex: Security & Observability for AI Agent Execution
Skil
Vendor-neutral security, verification, and enforcement framework for AI agent skills
AI Infra Security
Vulnerability research on AI infrastructure: LLM gateways, MCP transports, and model-serving stacks. Root cause, reproduction, and detection.
Capbroker
A local, self-hosted capability broker for AI agents -scoped, signed, expiring capability tokens instead of raw API credentials.
Attack Matrix
ATT&CK-style MCP attack matrix: tactics x techniques mapped to MITRE ATLAS, OWASP MCP Top 10 (2025) & SAFE-MCP, with detection notes and cited public research. A reference, not a…
Mcpsum
Stop MCP servers from rug-pulling your AI agent. A lockfile + runtime reference monitor for MCP tools.
Immune
Immune — the agent security loop that turns every attack into an antibody. MCP firewall → replayable breach trace in ClickHouse → synthesized countermeasure → replay-proven → Semg…
Scrutiny
Static and semantic security scanner for MCP servers: unpinned packages, rug pulls, tool poisoning, shadowing and toxic flows. Trained model, SARIF output, zero dependencies, full…
Vaast MCP
Connect VAAST to your AI assistants
Strathon
The open-source AI agent firewall — write a CEL policy, block the tool call before it executes. In-process, sub-millisecond. OWASP + EU AI Act mapped.
Brokenbazaar
Deliberately vulnerable AI agent marketplace — the full Product Security lifecycle: threat model → attack → defend → regression test → detect. MCP + Agent Skills supply chain.
Security Auditor
It's a Python-based CLI security scanner built specifically to audit MCP (Model Context Protocol) servers — the integration layer that connects AI assistants like Claude to extern…
Mcd Guard
麦麦护栏：把麦当劳账号交给 AI 之前，先装个刹车。本地 MCP 安全层：下单等写操作需本机确认、消费上限、只读模式、审计日志，Token 只在本机。Human-in-the-loop guard proxy for McDonald's China MCP.
Mcptrustchecker
Local-first, deterministic security scanner for Model Context Protocol (MCP) servers. Cross-tool toxic-flow analysis, Unicode-smuggling decode, prompt-injection & supply-chain det…
Adversarial Suite
🔒 Simulate adversarial behaviors to test and strengthen MCP defenses without real exploitation or risk, ensuring robust security evaluations.
Shield
AI-powered security scanner for MCP servers. Tests against the OWASP MCP Top 10 — prompt injection, tool poisoning, SSRF, and more — using a LangGraph attack agent.
Siege
Runtime red-team harness for live MCP servers. Attacks as real roles to find authz/RBAC bypasses a static manifest scan can't see. The offense leg of an agent-governance suite (Wa…
Gateway
Local-first firewall and black-box run reporter for MCP tool calls: block poisoned tools, detect drift, rate-limit agents, and audit runs.
TrustMCP
Zero Trust MCP server protected with SPIFFE/SPIRE, strict mTLS, and identity-based RBAC.
Warden
Drop-in security proxy for MCP — policy, tamper-evident audit, human approval, prompt-injection & secret-exfil defense. One line of config, zero code.
Toolgate
Capability gating for MCP agents, and the measurements that argue for it
Breaker
Replay-verified MCP security testing and deterministic policy hardening for tool-using AI agents
Guardbench
Check your MCP servers for tool poisoning and rug pulls, and benchmark MCP security tools under verifiable conditions.
Custody
Provenance-tracked shared memory for AI agent teams. A deterministic Guard runs multi-hop graph queries in FalkorDB to block injected actions, and a Warden agent investigates inci…
Shield
Semantic LLM firewall for the Model Context Protocol. A local proxy that classifies every untrusted tool result, tool description and server request with an 86M encoder, pins tool…
CHIMERA
Chained Harm & Injection Measurement for Evaluating Risks in Agents — a red-team research framework that tests hosted models, agentic apps, RAG pipelines, MCP servers, AI coding a…
Awesome Agent Supply Chain Security
A curated list of tools, research, standards and datasets for the security of AI agent extensions — skills, MCP servers, plugins and rules files.
Invaris Labs Website
Website for Invaris Labs: security infrastructure for autonomous AI agents. Presents AgentSec (adversarial security testing) and AgentAuth (agent identity and delegated authorizat…
Magate
生成AIを社外のMCPサーバやAPIにつなぐ前に、ツール定義・API仕様書の汚染・差し替え・乗っ取りや過剰な権限を審査し、許可／条件付き許可／不可を判定するツール
Two Key
Authorization and execution control for AI agents — independent policy and judge approval before consequential tool actions execute. MCP gateway, capability tokens, auditability,…
Agentgate
ai security cli developer-tools agent firewall policy-as-code python anthropic claude-code
Db2 MCP Server
Secure, observable Java 21 MCP server for governed read-only AI access to IBM Db2.
Trust
Preflight security scanner for MCP servers — map capabilities, scan install command, code, dependencies & metadata, and get an evidence-based APPROVE/BLOCK decision. CLI + GitHub…
Doctor
Scan, Score, Install MCP Servers with confidence. The missing toolkit for the MCP ecosystem.
Hackers Terminal Setup
One-script macOS/Linux security terminal setup — zsh, Powerlevel10k, tmux, and pentest/bug-bounty/cloud/AI-LLM/MCP tooling, installed from an interactive menu
Agent Action Gate
Deterministic policy gates for AI agent tool calls: allow, deny, or require human approval.
MCPShield
Local security scanner, hardening checks and optional stdio policy guard for Model Context Protocol (MCP) servers. Source beta by Nullvora.