FlagThis — Daily Cybersecurity Intelligence Briefing

FILTERING BY: CLEAR FILTER

AutoGen, LangGraph, and CrewAI: Self-Propagating Prompt Payloads Target Agent Frameworks

Self-propagating prompt payloads targeting Microsoft AutoGen, LangChain LangGraph, and CrewAI enable autonomous agent-to-agent malware propagation without host code execution vulnerabilities. Malicious natural-language instructions embedded in agent inputs, tool outputs, shared memory, or poisoned templates execute within LLM reasoning loops and replicate across inter-agent message queues, state graphs, and delegation chains. Attack vectors include indirect prompt injection, recursive delegation commands, and vector database memory poisoning. Impacts include unauthorized tool and code execution, lateral movement across agent swarms, data exfiltration, orchestration logic compromise, and supply chain contamination via malicious agent skills. Remediation requires strict I/O sanitization, least-privilege tool access, sandboxed execution, and runtime monitoring of shared state.

TWINLOOT: Microsoft 365 Weaponized as C2 Platform for Enterprise-Native Implants

TWINLOOT is a Python-based enterprise-native implant that weaponizes Microsoft 365 services for command-and-control. Attackers use external Microsoft Teams tenant messaging to impersonate IT support and deliver a PowerShell execution chain. The implant leverages SharePoint dead-drops for C2 and exfiltration, abuses Teams TURN relays, and controls headless Microsoft Edge via Chrome DevTools Protocol to blend traffic with legitimate browser activity. It harvests credentials via fake Windows lock-screen prompts and enables lateral movement through an integrated SOCKS5 proxy. Observed across at least three independent threat actors, this TTP cluster bypasses perimeter controls by routing malicious traffic through trusted M365 infrastructure.


LINK COPIED TO CLIPBOARD