articleDOAJ (DOAJ: Directory of Open Access Journals)Apr 1, 2026GREEN OA

From prompt injections to protocol exploits: Threats in LLM-powered AI agents workflows

MAMohamed Amine FerragNTNorbert TihanyiDHDjallel HamoudaLMLA MaglarasALAbderrahmane Lakas

United Arab Emirates University · Eötvös Loránd University · +4 more institutions

Indexed indoaj

Abstract

No abstract available for this paper.

Citation impact

12
total citations
FWCI
214.62
Percentile
100%
References
0
Citations per year

Authors

6

Topics & keywords

Topics
Keywords
  • Exploit
  • Workflow
  • Vulnerability (computing)
  • Protocol (science)
  • Fuzz testing
  • Bridging (networking)
  • Cryptographic protocol
  • Threat model
No related works found for this paper.