guardbee.ai

@guardbee/guard-core

Shared detectors for GuardBee MCP packages — checksum-validated PII masking, prompt injection patterns, and lethal-trifecta tool classification

@guardbee/mcp-a2a-auditor

MCP server that scans Agent2Agent (A2A) protocol implementation code for anti-patterns found in the reference SDKs and sample agents: unauthenticated push-notification webhook fetches (SSRF), Agent Cards with no authentication requirement, and credentials

@guardbee/mcp-agent-graph-auditor

MCP server that builds a reachability graph across multi-agent orchestration configs (LangGraph, CrewAI, AutoGen/ag2) to find transitive excessive agency — an agent with no dangerous tool of its own that can still reach one through delegation or group mem

@guardbee/mcp-ai-code-scanner

MCP server that scans code for insecure LLM/AI integration patterns — client-exposed API keys, unsafe handling of model output, excessive agent tool permissions, and PII sent to third-party LLMs

@guardbee/mcp-audit-gap-auditor

MCP server that audits MCP server source for OWASP MCP08:2025 (Lack of Audit and Telemetry) — missing tool-call logging, raw args/results in logs, disabled audit sinks, and tool errors swallowed without a record

@guardbee/mcp-compliance-checker

MCP server that checks KVKK, GDPR, and CCPA compliance for websites via the GuardBee API

@guardbee/mcp-config-auditor

MCP server that audits installed MCP client configs (Cursor, Claude Desktop, Windsurf, VS Code) and agent SKILL.md files for unpinned packages, secrets, wildcard auto-approve, unauthenticated remote endpoints, typosquats, unrestricted skill tools, cross-s

@guardbee/mcp-context-oversharing-auditor

MCP server that audits agent/MCP source for OWASP MCP10:2025 context oversharing — unrestricted session/memory dumps, cross-user shared context, system-prompt leakage via tools, and unscoped vector recalls

@guardbee/mcp-db-gateway

KVKK/GDPR-compliant MCP server gateway between LLMs and databases

@guardbee/mcp-dependency-auditor

OSV CVE checks for npm and pip manifests. Single-package lookups also cover PyPI, crates.io, Maven, Go, and RubyGems.

@guardbee/mcp-dns-intelligence

MCP server that enumerates DNS records, detects misconfigurations, and identifies dangling subdomains

@guardbee/mcp-document-scanner

MCP server and CLI that finds PII and credentials inside PDFs, screenshots and scanned documents — PDF text extraction plus optional OCR (Turkish and English) — before they reach an LLM

@guardbee/mcp-elicitation-auditor

MCP server that scans MCP server code for elicitation anti-patterns in the 2026-07-28 spec: form mode asking for secrets, URL mode pointed straight at a third-party authorize endpoint, and credentials or PII embedded in the elicitation URL

@guardbee/mcp-llm-redteam

MCP server that red-teams a live LLM endpoint or chatbot with canary-based jailbreak/extraction/obfuscation probes to test whether known bypass techniques defeat its guardrails

@guardbee/mcp-memory-poisoning-scanner

MCP server that scans agent code for memory-poisoning risks — untrusted input written into persistent, cross-session memory (MemGPT-style archival/core memory, or a vector store used as a long-term knowledge base) that later gets recalled and trusted as c

@guardbee/mcp-model-scanner

MCP server that scans ML model files (PyTorch, pickle, safetensors, Keras/H5, ONNX) for supply-chain risks — malicious pickle deserialization gadgets, disguised/malformed safetensors headers, Keras Lambda-layer RCE, and ONNX external-data path traversal

@guardbee/mcp-oauth-auditor

MCP server that scans an MCP server's own authorization code for OAuth 2.1 anti-patterns named in the MCP spec's Security Considerations: token passthrough, missing audience validation, OAuth/OIDC discovery SSRF, missing PKCE, loose redirect_uri validatio

@guardbee/mcp-owasp-scan

Unified OWASP MCP Top 10 scanner — one CLI/MCP that orchestrates GuardBee auditors against a path or a live tools/list, maps findings to MCP01–MCP10, and grades A–F

@guardbee/mcp-prompt-injection-scanner

MCP server that scans documents, scraped pages, and RAG content for indirect prompt injection payloads — instruction-override phrases, spoofed chat-template roles, hidden/invisible text, content addressing the AI directly, and data-exfiltration attempts

@guardbee/mcp-prompt-leak-scanner

MCP server and reverse proxy that catches leaked credentials and PII (API keys, TC Kimlik No, credit cards, IBANs) in outbound LLM prompts before they leave your application

@guardbee/mcp-rug-pull-detector

MCP server that connects to another MCP server (stdio or Streamable HTTP), baselines its tools/list response, and detects 'rug pulls' — a tool silently changing description, schema, or annotations after it was already approved

@guardbee/mcp-secret-scanner

MCP server that scans files and directories for exposed secrets, API keys, and credentials

@guardbee/mcp-security-proxy

MCP security gateway: several MCP servers behind one policy, with toxic-flow (lethal trifecta) blocking, approvals, PII masking and a hash-chained audit log

@guardbee/mcp-server-auditor

MCP server that audits other MCP servers' tool definitions for insecure patterns — excessive-agency tool names, shell/eval/SQL/SSRF sinks fed by raw tool input, overly permissive schemas, hardcoded secrets, and wildcard CORS

@guardbee/mcp-slopsquat-scanner

MCP server that checks every dependency declared in package.json/requirements.txt/pyproject.toml against the real npm/PyPI registry -- flags names that don't exist at all (likely LLM-hallucinated dependencies an attacker may have pre-registered, i.e. slop