← ProjectsSentinel
Case studySECURITYPERSONAL
Sentinel
A design-stage project for applying existing security policies to coding-agent actions.
Problem
Coding agents can take actions through tools, while existing scanners and policy checks usually report findings rather than decide whether an agent should be allowed to continue.
Architecture
The design is to inspect an agent’s tool request, check it against configured policy and existing security-tool findings, then allow or deny the action. The goal is to use those tools as inputs rather than build another scanner. Sentinel is still in design; this flow has not been implemented.
What I Built
A product design for policy checks around coding-agent tool calls.
An initial scope for connecting existing security findings to allow/deny decisions.
Outcome
Sentinel is still in the design stage. There is no working CLI or enforcement service yet.
Personal · Design stage