signal_tag · 3_broadcasts
#agent-security
// 3 transmissions tagged with #agent-security
TX_867236· Engineering
47,000 downloads of backdoored LiteLLM expose supply-chain weakness
A malicious LiteLLM build on PyPI was downloaded 47,000 times in three hours, highlighting a systemic supply-chain weakness in agent frameworks. Five concrete hardening steps, backed by OWASP data, can mitigate this risk [DevTo].
TX_007298· Engineering
Connecting an MCP server gives your agent hands, and a stranger a way in
Linking an MCP server lets an LLM‑driven agent call APIs and read files, but it also opens a two‑fold attack surface: malicious commands and poisoned data. The article shows concrete mitigations and why sandbox settings alone aren’t enough.
TX_501696· Crypto & Web3
Cord Protocol launches post-quantum identity SDK
Cord Protocol, an open source post-quantum cryptographic identity SDK, issues verifiable credentials for AI agents, addressing spoofing and prompt injection risks [Dev.to].