GuardBee/guardbee-mcp

A family of 25 independent, MIT-licensed MCP servers for AI/LLM and application security scanning — secrets, CVE dependency auditing (OSV), TLS/DNS inspection, prompt-injection and insecure-tool-definition scanners, an Agent2Agent (A2A) protocol auditor, a slopsquatting detector, and an Unbounded Consumption ("denial of wallet") auditor. Each is a standalone npm package under `@guardbee/mcp-*` with `scan_text`/`scan_file`/`scan_directory`/`list_patterns` MCP tools and SARIF output. Docs: https://guardbee.ai.

Project

Browse this MCP server in the TensorBlock MCP Index.