write
Source code
Repositories, pull requests, issues, and workflows.
3 modeled gaps
AI Agent Blast-Radius Mapper
See what an AI agent can reach—and what happens if its controls fail.
Model tools, data, credentials, action authority, approval gates, validation, logging, and recovery. The mapper produces a live authority view and a prioritized control report.
Live authority model
Current blast radius
Reachable systems and the agent's highest modeled authority.
4 connected systems
AI agent
MCP-connected assistant
send
Mailboxes, drafts, messages, and recipients.
4 modeled gaps
read
Customer data
Profiles, orders, cases, and account records.
No modeled gaps
execute
Deployment
Build systems, environments, releases, and production.
4 modeled gaps
This mapper is a planning aid based on the authority and controls you select. It is not a certification, exhaustive threat model, or substitute for validating the deployed system.
Method
Authority determines impact.
Map reachable systems
List each system the agent can reach and distinguish read, write, send, and execute authority.
Model deterministic boundaries
Count controls only when code, policy, credentials, or human approval enforce them outside the model.
Prioritize control gaps
Focus first on paths that combine sensitive systems, external actions, production authority, and weak recovery.