Prompt injection through delegation chains
Recent work on multi-step compromise, tool misuse, and indirect prompt injection in agent workflows.
Research index
Multi-agent security research.
A focused index for recent papers, topic groupings, concepts, and trend signals related to multi-agent security.
Recent publications
Some of the most recent papers
Trust, identity, and cross-agent verification
Recent work on dynamic trust scoring, role separation, and verification in agent-to-agent interaction.
Memory poisoning and long-horizon manipulation
Recent work connecting retrieval, persistent memory, and long-horizon manipulation in agent systems.
Research trends
Where the literature is moving.
Prompt injection, orchestration risk
Trust scoring, evaluation benchmarks
Agent identity, long-horizon governance