APEX: Active Protection at Execution Boundaries for LLM Agents

Indirect prompt injection IPI hides adversarial instructions in content that large language model LLM agents read at runtime. As agents compose heterogeneous capability units, including Tools, MCP servers, and Skills, the carriers of injection multiply, and defenses built to recognize attack...

Remote Code Execution (RCE)

gerbv is vulnerable to remote code execution. The vulnerability exists due to a boundary error when processing untrusted input in the drill format T-code tool number functionality...

lcdproc lcdd 0.x4.x - Multiple Vulnerabilities

lcdproc lcdd 0.x4.x - Multiple Vulnerabilities source: https://www.securityfocus.com/bid/10085/info LCDproc Server LCDd has been reported to be prone to multiple remote vulnerabilities. The first issue is reported to exist in the parseallclientmessages function of parse.c, and is due to a lack of...