Question 1

What is the policy playground?

Accepted Answer

The Aegis Mesh playground is a live, interactive policy environment for kernel-level AI agent governance. Pick a scenario, see the agent tools the scenario exposes, see the enforcement policies attached to it, fire actions from the right rail, and watch each decision stream into the architecture canvas and the forensic ledger in real time. Latency, decision counts, and allow/escalate/block tallies are reported per session. Every entry is sealed into a SHA-256 hash-chained ledger you can export as JSON.

Question 2

What policies can I write?

Accepted Answer

Each scenario ships with enforcement policies authored in Rego (Open Policy Agent). Policies declare which tools, hosts, and exec calls each agent role can make. The playground exposes the full policy list per scenario in the left rail; firing an action evaluates the matching tool call against those policies and returns ALLOW, ESCALATE, or DENY. The full bundle is the same Rego format used in production deployments. What you see here is the same policy engine the proxy and BPF-LSM enforcer run against in the live demo.

Question 3

Where do I see decisions?

Accepted Answer

Decisions stream live over Server-Sent Events (SSE) into three surfaces. The architecture canvas in the centre animates each step of the request: proxy intercept, semantic firewall, policy engine, kernel enforcement. The Enforcement Metrics panel on the right tracks latency and counts. The Forensic Ledger panel at the bottom tabulates every decision with a Time, Tool Call, Decision, Reason, Component, and Hash column. Filterable by Allow / Escalate / Deny / Hibernate, exportable as JSON.

Aegis Mesh. Interactive Policy Playground.

Live Policy Playground

Security Champion