Security MCP servers for scanning, threat intelligence and audits: let an AI assistant check code, dependencies and infrastructure for risks. 3,540 servers in this category.
Provides anonymous, zk-proof gated web egress for AI agents through a local HTTPS proxy or MCP server.
Provides a visual and source-grounded guide to understanding agentic AI architectures and learning how to secure them.
Supplies an AI client with secure-coding prompts, agents, and skills to guide GitHub Copilot toward safer Java, Node.js, C#, and Python code.
Provides a secure, encrypted vault and fixed-operation broker to manage credentials for AI agents without exposing raw secrets.
Analyzes README files for portability issues and command risks using static analysis and provides reviewable verification plans.
An agentic AI framework that automates the ten-phase enterprise vulnerability assessment lifecycle using progressive context chaining.
Decompile Android APKs and inspect iOS IPA metadata to recover Java source code and resources for local analysis via AI agents.
Lets an AI client query an obligation atlas and check AI governance files against global regulatory regimes.
Acts as a security firewall and sandbox proxy to protect AI agents from malicious tool misuse, prompt injection, and credential exfiltration.
Acts as a security proxy for MCP servers to redact secrets, prevent prompt injection, and audit tool usage.
Provides secure, read-only web search and fetching capabilities with SSRF protection, DNS rebinding guards, and outbound data redaction.
Provides a secure proxy layer for AI agents to interact with computers, SSH servers, browsers, and cloud storage through human-in-the-loop approvals.
Performs pre-merge security scans on pull requests for malicious packages, MCP servers, GitHub Actions, and hidden Unicode in agent instruction files.
Provides OAuth authentication for stdio-only MCP clients by bridging them to remote servers using PKCE, token refresh, and secure token storage.
Hosts MCP App views in a web page using a sandboxed proxy on a second origin to protect the host's origin and security.
Simulates a safety checkpoint for voice requests by implementing policy screening, confirmation challenges, and tamper-evident auditing via MCP.
Acts as a governed proxy that classifies, decides, and records every MCP tool call before forwarding it to the upstream server.
Scans repositories for malicious hooks, MCP servers, VS Code tasks, and npm scripts before they are executed by a user or coding agent.
Protects Claude Code sessions by scanning remote MCP servers for tool poisoning and checking package installs against malware advisories.
Performs static security scanning and live metadata probing to detect prompt injection, tool poisoning, and secret leakage in MCP servers and agent skills.
