Inkog

Inkog is an MCP (Model Context Protocol) server focused on security auditing for AI agent and MCP server development. It exposes tools to scan for vulnerabilities—including logic bugs, prompt injections, missing guardrails, hardcoded secrets, and compliance gaps—across agent code and MCP tool integrations, supporting both code-first and no-code agent frameworks. Integrations include CI/CD systems (e.g., GitHub Actions), major agent frameworks (like LangChain, AutoGen, OpenAI Agents, and more), and direct scans from compatible MCP clients such as Claude, Cursor, and Windsurf. Designed for AI engineers, security professionals, and teams building or auditing AI agents, Inkog helps verify tool and agent governance, ensure compliance (e.g., EU AI Act, NIST AI RMF), and reduce risk before deployment.