Architecture Diagrams
Technical diagrams showing how AIM integrates with AI agents and provides security.
AIM Architecture Overview
How AIM integrates with AI agents and MCP servers
AI Agents
LangChain, CrewAI, AutoGen, and custom agents register with cryptographic identity
AIM Platform
Central security layer handling identity, access control, trust scoring, and audit logging
MCP Servers
Attested servers verified before connection, unverified servers automatically blocked
Security Flow
How every agent action is verified and authorized
Agent Request
Verify Signature
Check Capability
Trust Score
Execute/Block
8-Factor Trust Scoring
Real-time behavioral analysis algorithm
Identity & Behavioral
Signature verification and action consistency
Resource & Temporal
Access patterns and timing analysis
MCP & Policy
Server trust and compliance checking
Error & History
Error rates and historical behavior
MCP Server Attestation
Cryptographic verification before agent connection
Registration
MCP servers register their capabilities, tools, and access permissions with AIM
Verification
AIM creates cryptographic attestation with configuration hash and Ed25519 signature
Drift Detection
Continuous monitoring catches unauthorized configuration changes
Supply Chain Visibility
Complete dependency graph of agents to MCP servers
Agent Inventory
Track all registered agents with trust scores and verification status
Dependency Mapping
Visualize which agents connect to which MCP servers
Risk Assessment
Identify unattested servers and potential supply chain risks
MCP Asset Management
Complete inventory of MCP servers, owners, and capabilities
Server Inventory
Track every MCP server in your organization
Ownership
Know who registered each server and which team owns it
Capabilities
Understand what each MCP can access (read/write, APIs)
Shadow IT Detection
Identify and block unattested MCP servers
All diagrams available in SVG format for use in presentations and documentation.