Practitioner · 15–25 min
Agent Permission Lab
Paste an agent tool configuration JSON — including MCP schemas — and analyze it for permission scope problems, missing approval gates, and execution identity risks. 15 security rules across 6 risk categories.
Learning objectives
- Understand the three permission scope levels — read-only, write-broad, admin — and when each is dangerous
- Identify side-effecting actions (email send, record modification, code execution) that need approval gates
- Recognize when an agent's execution identity (service account, system admin) expands blast radius
- Read MCP tool schemas and spot ambiguous descriptions that hide capability scope
Risk categories
Permission scope
read-only vs write-broad vs admin
Side effects
email, record mod, privilege change, code exec
Approval gates
side-effecting actions without confirmation
Execution identity
unknown, service account, system admin
MCP quality
ambiguous descriptions, unbounded arguments
Framework
OWASP LLM06 — Excessive Agency
Reading materials
AIPSA Handbook · Ch 6
Chapter 6 — Agentic Permissions
Tool-call security, delegated authority, approval gates, side-effect containment, MCP surface analysis, sandboxing, and action logging for AI agents.
4.9 MB
AIPSA Field Guide · Ch 5 · Ch 5
Agent Security
Delegated authority, tool calls, MCP-style tool surfaces, approvals, side effects, action logging, sandboxing, and agentic workflow governance.
~2 MB
Mythos Report · Ch 9 · Ch 9
Excessive Agency Is the New Overprivileged Service Account
Drawing the direct line from classic least-privilege failures to agentic AI: why scope, approval gates, and blast radius matter more as agents gain capabilities.
~1 MB
Interactive tool