SECURE KNOWLEDGE QUERY MANIPULATION LANGUAGE:
     A SECURITY INFRASTRUCTURE FOR AGENT COMMUNICATION LANGUAGES
 
      Muhammad Rabi, Tim Finin, Alan Sherman, and Yannis Labrou
	     Computer Science and Electrical Engineering
	       University of Maryland  Baltimore County
		    Baltimore, Maryland 21228-5398


		      -- draft May 1998 draft --

Despite the security and privacy concerns that agents could encounter
whenever they cross multiple administrative domains, the agent
communication languages standards lack the necessary constructs that
enables the secure cooperation among software agents.  We propose
Secure Knowledge Query Manipulation Language (SKQML) as a security
infrastructure for KQML-speaking agents. SKQML enables KQML-speaking
agents to authenticate one another, implement specific security
policies based on authorization schemes, and whenever needed ensure
the privacy of the messages exchanged.

SKQML is simple, extensible, and at a level appropriate for
intelligent communicating agents. In addition, SKQML employs
public-key cryptographic standards and it provides security mechanisms
as an integral part of the communication language.  This paper details
the synthesis of public key certificate standards and agent
communication languages to achieve an infrastructure that meets the
security needs of cooperating agents with detailed examples using a
partial prototype implementation of this infrastructure.