Search

The security layer for AI agents — detect threats across 9 platforms with 316 rules, block them at runtime, guide remediation. Continuous protection.

189 security tests for AI agent systems - MCP, A2A, L402 wire-protocol testing, 20 enterprise platform adapters, GTG-1002 APT simulation

AI Agent risk scanner — detect security risks in skills, MCP servers & plugins. 29 rules mapped to OWASP Top 10 LLM, MITRE ATLAS & CWE standards. Offline, open source.

Offline security scanner for MCP servers: static config audit + source-code SAST (JS/TS/Python/Go) + opt-in live probing, with policy gates, tool-poisoning detection, secret redaction and SARIF/JUnit/CycloneDX output for CI. One Go binary, no LLM.

Open-source security scanner and policy gate for Model Context Protocol (MCP) servers.

ESLint plugin for Model Context Protocol (MCP) SDK security — catches tools registered without an input schema, so handlers never run on unvalidated client-supplied arguments.

Runtime security for AI agents. In-process, zero dependencies, Apache 2.0.

AI-layer security audit — scan system prompts, MCP configs & tool descriptions for prompt injection and toxic permissions; diff-driven prompt regression gate. AI 层安全审计与门禁：提示注入扫描 / MCP 安全 / prompt 回归测试

Security scanner for MCP servers. Find vulnerabilities before your AI agent does.

MCP is being adopted rapidly. Security guidance is lagging behind. This checklist gives security engineers, platform teams, and technical leaders a clear, actionable baseline for securing MCP deployments , whether you're shipping an internal tool or a customer-facing AI agent.

Security co-pilot for AI agents. Scan for vulnerabilities, verify governance, and generate compliance reports during development.

Lighthouse for AI agents. Security scanner with 25 prompt defense vectors (12 LLM-era + 13 agent-era, OWASP ASI01-ASI09), PII protection, AI visibility scoring.

Open-source AI security verification — model artifacts, live endpoints, MCP servers and recorded agent traces. One rule engine for your laptop, CI and production, mapped to OWASP/MITRE ATLAS/NIST. Deterministic evidence, a measured verdict, and an explicit "could not tell". Apache-2.0.

The Guardian Protocol - MCP-based security monitoring for agentic AI systems

Full-spectrum wireless security MCP server — WiFi (WPA/WPA2/WPA3, KRACK, Dragonblood, PMKID, Evil Twin), Bluetooth (BlueBorne, KNOB, BLE), RF/SDR (RollJam, sub-GHz), NFC/RFID (Mifare, EMV relay), Cellular/5G (IMSI, SS7), IoT (Zigbee, Z-Wave, LoRa, MQTT),

Free security assessment for your OpenClaw 🦞 environment. Scans gateway config, tool permissions, MCP servers, plugins, and chained attack paths.

Prompt injection + PII detection MCP server for AI agents. 22 signatures, 7 languages, ~23ms latency. Signed GDPR Art.30 audit reports.

Open-source security scanner for Model Context Protocol (MCP) servers. Audits Claude Desktop, VS Code, Cursor, Windsurf, and 16+ AI tools for secrets, prompt injection, supply-chain risks, and 17+ security checks.

Check anything before your agent trusts it; verify x402 payments and deliveries. Offline verdicts.

Audits the MCP servers configured on a machine for supply-chain risk: unpinned versions, inline secrets, missing provenance, install scripts, maintenance signals