tool-poisoning

Here are 8 public repositories matching this topic...

sahiloj / MCPScan

Offensive MCP server auditor — detects tool poisoning, credential leaks, RCE vectors, SSRF, session hijacking, and supply chain vulnerabilities across stdio, HTTP, and SSE transports.

open-source security mcp supply-chain offensive-security vulnerability-scanner ai-security llm-security model-context-protocol mcp-security tool-poisoning mcpscan

Updated Mar 12, 2026
TypeScript

elliotllliu / agent-shield

Star

🛡️ Security scanner for AI agents, MCP servers & plugins — 30 rules, AST taint tracking, cross-file analysis, kill chain detection. Free & open source alternative to Snyk Agent Scan.

mcp static-analysis security-scanner dify ai-security supply-chain-security ai-agent prompt-injection model-context-protocol mcp-server agent-security mcp-security openclaw tool-poisoning

Updated Mar 16, 2026
TypeScript

deconvolute-labs / deconvolute

Star

Real-time security layer protecting AI Agents from Confused Deputy attacks, malicious MCP payloads, and Indirect Prompt Injection.

security ai mcp firewall dns-rebinding ai-security prompt-injection llm-security rug-pull rag-pipeline agentic-ai model-context-protocol mcp-client mcp-security tool-poisoning

Updated Mar 11, 2026
Python

deconvolute-labs / mcp-deconvolute-demo

Star

Live PoC: MCP attacks that compromise AI agents mid-session and how to block them in a few lines of code.

python dns mcp cybersecurity ai-agents dns-rebinding ai-security security-demo llm prompt-injection llm-security rug-pull agentic-ai model-context-protocol mcp-security ai-agents-mcp tool-poisoning

Updated Mar 11, 2026
Python

TimesAndPlaces / mcp-preflight

Star

Local static scanner for MCP setup, config, prompts, and workflow trust.

cli mcp vscode-extension ai-security local-first supply-chain-security prompt-injection model-context-protocol tool-poisoning

Updated Mar 9, 2026
TypeScript

panguard-ai / panguard-ai

Star

Open-source security platform for AI agents -- audits skills before install, monitors 24/7, shares threat intelligence across all users. | AI Agent 開源安全平台 -- 安裝前審計 skill、24/7 即時監控、社群共享威脅情報。