Inter-Agent Communication

Multi-agent systems pass natural-language messages between agents, and each message is a potential injection vector. This type covers insecure or spoofable agent-to-agent channels and the attacks that ride on them — OWASP ASI07.

OWASP Agentic Top 10: ASI07 Insecure Inter-Agent Communication

Other agent threat types

Prompt InjectionMCP Tool PoisoningTool MisuseMemory PoisoningAgent WormsPrivilege AbuseUnsafe Code ExecutionData ExfiltrationProtocol VulnerabilitiesFramework VulnerabilitiesSupply-Chain CompromisesGoal HijackingIdentity SpoofingInter-Agent CommunicationCascading FailuresHuman Trust ExploitationRogue AgentsResource Exhaustion

Showing 1–1 of 1 threats, newest first

Undefended Inter-Agent Channels Enable Tool/Memory Poisoning and Prompt Injection in Multi-Agent Pipelines

mediumAgentInter-Agent Comms
multi-agenttool-poisoningprompt-injectiondefense-in-depthresearchinformation-bottleneckbenchmarkingprovider-side-filter-dependenceASI05 · Unsafe Code ExecutionAML.T0051AML.T0054Surface: Inter Agent CommsPropagation: Single Hop

This is a research paper (not an active exploit) demonstrating that multi-agent LLM pipelines composed of individually safe models are not safe by default, because the hops between planner, worker, verifier, and synthesizer agents are unmonitored channels an adversary can use to smuggle instructions. The authors show that apparent 'zero attack success' in undefended pipelines was largely an artifact of cloud-provider server-side filtering rather than genuine architectural safety, and propose a training-free gating defense (ChannelGuard) that meaningfully reduces tool-poisoning and prompt-injection success. Severity is moderate: this is a measurement/defense study highlighting a real but already partially-known class of risk, not a novel zero-day.

Updated Jul 23, 2026