Security Advisories
Security research focused on Model Context Protocol (MCP) implementations and AI-assisted development tools.
Published Advisories
-
Kluster's Verify MCP Server Exposes Users to Credit Exhaustion
Kluster AI’s verify-mcp server trusts any browser session that can reach its /stream endpoint.
-
Neo4j MCP Cypher Server Vulnerable to Database Takeover Via DNS Rebinding
A DNS rebinding vulnerability in the Neo4j MCP Cypher Server allows remote attackers to execute arbitrary Cypher queries against a user's database, leading to potential data theft, modification, and full...
-
Vet MCP Server SSE Transport DNS Rebinding Vulnerability
SafeDep Vet MCP Server is vulnerable to DNS rebinding attacks allowing malicious websites to bypass Same-Origin Policy and exfiltrate scan database contents through unauthorized MCP tool invocations.
-
Amp AI Agent Allows API Key Exfiltration Via Prompt Injection
A prompt injection vulnerability in Amp's CLI and extensions allows attackers to exfiltrate environment variables including API keys via DNS queries without user consent. Amp declined to address the issue,...
-
Kilo Code AI Agent Exposes Users to Supply Chain Attack Via Prompt Injection
A prompt injection vulnerability in the Kilo Code AI agent allows attackers to modify application settings, whitelist arbitrary commands, and execute automated supply chain attacks without user interaction.
-
Coder's Agent API Exposes User Chat History Via DNS Rebinding Attack
A DNS rebinding vulnerability in Coder's Agent API allows remote attackers to exfiltrate a user's entire local message history, which may contain sensitive data like secret keys and intellectual property....
-
Unauthorized Crypto Transactions Enabled by thirdweb MCP Server
thirdweb MCP Server exposes unauthenticated SSE interface enabling unauthorized cryptocurrency transactions from victims' wallets.
-
Grafana MCP Server Exposes Unauthenticated SSE Interface Enabling Remote Dashboard Manipulation
Grafana MCP Server exposes unauthenticated SSE interface allowing network-level attackers to manipulate Grafana dashboards and access sensitive data.