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.
-
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 -
guardianshieldgithub.com/sparkvibe-io/guardianshieldupdated this year
Universal AI security layer - MCP server for code scanning, PII detection, prompt injection defense, secret detection, and audit logging
pypi Apache-2.0 -
MCP injection / prompt-poisoning / SSRF scanner. 30+ canonical rules covering the April 2026 Anthropic MCP RCE class, mcp-server-git CVE chain, DockerDash injection chain, OWASP LLM Top 10 + GenAI Red Team v1. Scans live MCP /tools/list endpoints OR pasted JSON. Free 5/day, Pro £29/mo for unlimited + signed safety reports auditors and procurement accept. By MEOK AI Labs.
pypi MIT License -
prompt-protectiongithub.com/mughalhere/prompt-protectionno update data
Scan prompts, tool definitions and model output for injection, and guard agent tool calls.
npm -
Local-first trust layer for AI coding agents: structural prompt-injection and action guards, secret scanning, semantic memory, code review, and 17 MCP tools. No model API required.
MIT ★ 1 -
PromptGuard Scan remote MCP documentation for prompt injection and LLM security testing.
NOASSERTION -
@prompt-anonymizer/mcpgithub.com/akazah/prompt-anonymizerupdated this year
A buddy check for PII before it reaches an LLM - consistent, reversible labels. MCP server (anonymize / deanonymize / scan tools) built on @prompt-anonymizer/core.
npm MIT -
MCP server: scan web pages, mails, PDFs and tool output for prompt injection before your agent reads them (17 languages).
MIT -
agentguard-mcpupdated this year
AgentGuard MCP server — scans Cursor tool responses for prompt injection and PII leaks
npm MIT -
@mukundakatta/prompt-injection-shield-mcpgithub.com/mukundakatta/prompt-injection-shield-mcpupdated this year
MCP server exposing prompt-injection-shield: scan untrusted text for prompt-injection signals, score risk, strip dangerous lines. Built for Claude Desktop, Cursor, Cline, Windsurf, and Zed.
npm MIT -
promptguard-mcp-servergithub.com/acebot712/promptguardupdated this year
PromptGuard MCP server — scan prompts, redact PII, and audit LLM SDK usage from any MCP client
pypi MIT -
mcp-sandbox-scaupdated this year
MCP server that scans repos and AI tools for prompt injections, exfiltration, and malicious behavior
npm MIT -
Stop prompt injection in AI inbox and chat agents: scan emails, pages and messages for instructions aimed at the AI, fence untrusted content for the prompt, and gate the tool calls an agent makes after reading it. CLI, library, MCP server. Zero dependencies.
npm MIT -
security-mcpgithub.com/yeodh10/security-mcpupdated this year
An MCP server exposing security tools (prompt-injection scan + CVE lookup/version-impact) that Claude calls directly. FastMCP, reuses cve-radar/prompt-guard logic.
-
malskannergithub.com/Octolabo/malskannerupdated this year
Scan a repo for hidden prompt-injection before your AI agent trusts it — CLI, MCP server & GitHub Action. Run: npx malskanner <repo>. 0 false positives across 5,620 files.
MIT ★ 5 -
@muhannad-hash/mcp-shieldgithub.com/muhannad-hash/mcp-shieldupdated this year
MCP server that scans other MCP servers for backdoors, prompt injection, exfiltration, and supply chain risks
npm MIT -
@safekeylab/mcp-servergithub.com/safekeylab/mcp-serverupdated this year
MCP (Model Context Protocol) server for SafeKeyLab AI security tools - PII scanning, prompt injection detection, and model security analysis
npm MIT -
MCP server and CLI that scan untrusted content for prompt-injection patterns before it reaches your agent - offline, explainable, cross-agent.
MIT ★ 1 -
adler-mcpgithub.com/commit3296/adlerupdated this year
Model Context Protocol server for the Adler OSINT username-search engine — exposes scan / list / doctor / history tools, registry + scan-history resources, and OSINT workflow prompts.
cargo MIT -
agent-immunegithub.com/denial-web/agent-immuneupdated this year
Adaptive threat intelligence for AI agent security — semantic memory, multi-turn escalation, output scanning, rate limiting, and prompt hardening.
pypi Apache-2.0 -
agent-auditgithub.com/ondraulehla/agent-auditupdated this year
Security auditor for AI agent configurations — scans MCP servers, Claude Code settings, agents, skills and hooks for risky permissions, inline secrets, unpinned supply chains and prompt-injection surfaces.
MIT -
simply-ops-guardgithub.com/nardovibecoding/simply-ops-guardupdated this year
Claude Code plugin - 49 hooks + 28 MCP tools for enforcement, security scanning, prompt injection defense, VPS ops, audit logging
AGPL-3.0 ★ 1 -
cheetah-firewall-mcpupdated this year
MCP server for agent safety: prompt-injection scan, URL reputation, on-chain agent trust, AML/sanctions screening and skill pre-install analysis, paid per-call over x402 (free tier included).
pypi MIT -
YUSAgithub.com/yusasecurity/yusano update data
Prompt-injection firewall for AI agents: scan untrusted text, fence it, gate tool calls.
mcpb -
MCP server exposing Bawbel Scanner as agent-callable tools. Scan MCP servers, skill files, and system prompts for AVE vulnerabilities mid-conversation.
pypi Apache-2.0 ★ 2 -
TollWarden — a payment security firewall for x402 micropayment traffic. Advisory scanning: PII/secret leaks, replay, overpayment, prompt-injection-triggered payments, counterparty reputation. Non-custodial. Includes an MCP server: npx tollwarden
npm BUSL-1.1 -
AI-DOS: Code Intelligence Platform — knowledge graphs, security scanning, and closed-loop AI development via MCP tools, sub-agent prompts, security rules, and CLI workflows.
npm MIT -
agent-security-mcpgithub.com/mdfifty50-boop/agent-security-mcpupdated this year
MCP server providing security scanning, prompt injection detection, secret leak detection, and agent permission auditing for AI agent workflows
npm -
Agent Security Scannerno update data
Scan AI agents for tool-calling vulnerabilities: prompt leaks, hijacking, injections, and more.
pypi -
Symbol-level npm dependency analysis: scan verdicts, native alternatives, migration prompts.
npm -
@iflow-mcp/agentsafe-tooltrustgithub.com/agentsafe-ai/tooltrust-scannerupdated this year
MCP server that scans other MCP servers for prompt injection, data exfiltration, and privilege escalation. Add to your .mcp.json and let your AI agent audit its own tools.
npm MIT -
paysafe-x402github.com/corbinallison/paysafeupdated this year
PaySafe — a payment security firewall for x402 micropayment traffic. Advisory scanning: PII/secret leaks, replay, overpayment, prompt-injection-triggered payments, counterparty reputation. Non-custodial. Includes an MCP server: npx paysafe-x402
npm BUSL-1.1 -
noatmark-mcpgithub.com/ylwl1997/noatmark-text-hygieneupdated this year
MCP server for text hygiene: sanitize text before feeding it to an LLM (strip invisible characters, flag prompt injection), scan for hidden text, and clean LLM formatting. Companion to noatmark.com.
npm MIT -
Secure MCP server — scans web content for prompt injection before it enters LLM context. Rust rewrite of webguard-mcp.
MIT -
MCP server: prompt-injection scan (PITAX) + deterministic code audit for AI-assisted coding. Local stdio, no model calls, no telemetry.
MIT -
Audit what your AI agents can actually do to your machine. Scans installed MCP servers, skills, hooks and agent configs for leaked credentials, hidden prompt injection, exfiltration and over-broad permissions. Offline, read-only, zero dependencies.
MIT -
MCP server for Power Strip Security: read web pages, scan content for prompt injection, and audit MCP servers and agent skills, paid per call with x402 (USDC on Base) from your own wallet, with a spending cap.
npm MIT -
Security scanner for third-party AI agent-skill files (SKILL.md manifests, hooks, bundled scripts). CLI, library, MCP server, and GitHub Action, all reading the same 10 rule packs, including cross-skill privilege chaining and prompt-injection detection. Zero-auth, zero-config npx or pip scan.
pypi Apache-2.0 ★ 3 -
MCP server: 7 site and email safety tools - SPF/DKIM/DMARC, AI/SEO audit, prompt-injection scan, RSS monitor.
pypi MIT -
Gadriel AI Security Harness for GitHub Copilot — local MCP server, prompts, and instructions for SAST, SCA/SBOM, secrets & OWASP LLM security scanning.
npm Apache-2.0 ★ 2 -
@clarxai/mcpactive
Clarx MCP server — score agent manifests, validate clarx-manifest.json, generate CI workflows, and (with a Clarx token) pull scan findings and remediation prompts into your IDE.
npm MIT -
uiscanner-mcpgithub.com/mishaovcharenko/uiscannerupdated this year
MCP server + CLI for uiscanner: scan any public URL for a design teardown (tokens + section structure + a Claude-tailored build prompt).
npm MIT -
redproof-mcpactive
MCP server for Redproof: security checks for apps built with AI tools. Scan a GitHub repo and read the fix prompts from Claude, Cursor or any MCP client.
npm MIT -
promptguard-mcpgithub.com/KrishOjha1810/promptguard-mcpupdated this year
Local-first MCP server that scans developer prompts for secrets, PII and token cost before they reach the LLM. TypeScript, with VS Code and browser extensions.
MIT ★ 1 -
MCP server that lets AI coding agents (Cursor, Claude Code, Windsurf) scan landing pages for AI-design-slop and pull a fix prompt.
npm MIT -
MacVM-MCPgithub.com/zixuantemp/MacVM-MCPupdated this year
MCP server for automated macOS malware analysis via SSH. 43 tools, 5 prompts, 5 resources, composite playbooks (triage, behavioral, app-bundle audit, IR scan).
MIT -
Declaw -- Secure AI Sandboxesno update data
Secure runtime for AI agents. Run AI agents and their tools in Firecracker microVMs with network policies, PII scanning, prompt injection defense, and audit logging. 7 tools: create_sandbox, run_command, read_file, write_file, list_files, kill_sandbox, list_sandboxes. Security presets (none / standard / strict) let you control guardrails per sandbox: restrict outbound traffic to specific domains, redact PII from command output, block prompt injections, and get full audit trails.
smithery -
wardloom-mcpactive
MCP server for Wardloom — trigger pre-deploy security scans, poll the ship-gate verdict, fetch remediation roadmaps and Dev/AI handoff prompts, and run zero-network local prechecks, all from your IDE's AI agent.
npm MIT -
A Model Context Protocol server with machine learning engineering tools: BPE token counting and cost estimates, BM25 search, paired bootstrap and McNemar evaluation comparison, JSON Schema validation, and prompt-injection and PII scanning.
MIT -
@sonnylabs/mcpgithub.com/sonnylabs/sonnylabsupdated this year
Model Context Protocol server for the Sonny Labs AI firewall. Run it inside your agentic client (Claude Desktop, Cursor, Claude Code) to scan prompts and outputs for injection / PII / toxicity, manage API keys, and scaffold the firewall into your codebase
npm Apache-2.0 -
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 -
LangfuseMCPgithub.com/DrishtantKaushal/LangfuseMCPupdated this year
Langfuse MCP server with built-in analytics. 34 tools — traces, observations, sessions, scores, prompts, datasets + accuracy metrics, failure detection, token percentiles, cost breakdowns, latency analysis, context breach scanning. Works with Claude Code, Cursor, Codex.
MIT ★ 2 -
55-tool MCP server for security scanning of MCP servers. Runtime inspection, AST-based SAST, config audit, dependency analysis, OWASP MCP Top 10 compliance. OAuth, TLS, fuzz testing, prompt injection, tool mutation detection. 100% local, zero external API calls.
MIT ★ 5 -
Catch the instructions hidden in a document before your agent follows them. Finds text a model reads but a person cannot see in PDF, DOCX and HTML. CLI, library and MCP server.
Apache-2.0 -
Security scanner for MCP servers, as a CLI, GitHub Action and MCP server: tool poisoning, prompt injection, secret leaks and excessive permissions, with SARIF output for CI.
npm Apache-2.0 -
agent-skill-scan-mcpgithub.com/rexcoleman/agent-skill-scan-mcpupdated this year
MCP server: Scan OpenClaw SKILL.md and MCP tool definitions for security vulnerabilities. 22 rules. For Claude Code.
MIT -
Free offline scan of your AI agent: what loads on every message, heartbeat costs, plaintext API keys and planted or bloated memory. CLI, agent skill and MCP server. Runs locally, sends nothing. MIT-0.
npm MIT-0 -
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 -
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 -
MCP Server for ClawGuard Shield — security scanning via Model Context Protocol.
pypi MIT ★ 1