Proof of compatibility
Install OpenShell and XMAN in a lab, verify rootless lifecycle behavior and document supported APIs.
OpenShell enforces what the agent may do inside its runtime. XMAN targets the surrounding human and operational process. Here is the integration hypothesis—and what must be proven before shipping.
Architecture fit
| Question | OpenShell | XMAN / Clawsolator direction |
|---|---|---|
| Can a process read, write, connect, or call a tool? | Enforce external runtime policies. | Coordinate compatible sandboxes and independently gate specified host actions. |
| Who can grant permission? | Gateway/policy approvals and operator administration. | Named requests, approval histories and operator review workflow. |
| What if an agent runs away? | Contain process/network activity according to policy. | Resource budgets, timeout and job scheduling controls under development/implementation review. |
| Can we rebuild evidence after a restart? | Runtime audit and policy records. | Durable request state, recovery pathways and attributable logs—verify per release. |
| Are they already integrated? | No. Future compatibility is being evaluated; no shipped OpenShell integration is claimed. | |

“If OpenShell already does the sandbox and the egress rules, why would I want XMAN?”

“Because an organization also needs operator-owned change management, recovery, accountability, agent assignment and resource plans. But we should integrate—not duplicate enforcement.”
Scope note: NVIDIA already documents strong external enforcement and auditing. Any claim that OpenShell lacks approval, rootless Podman, or network control would be inaccurate.
Install OpenShell and XMAN in a lab, verify rootless lifecycle behavior and document supported APIs.
Make explicit which system is authoritative for process, network, filesystem and human approval gates.
Correlate request IDs and OpenShell policy decisions without creating a second source of truth.
Test restart, denial, policy changes, offline operation and resource exhaustion, then publish redacted results.
Explore the evidence
Inspect the controls, verify the boundaries, and ask what happens after a failure—not merely before it.
Explore the security lattice →Read the architecture →CANONICAL TRUST BOUNDARY
Each registered agent uses the authenticated requester API through the trusted Gateway bridge. The XMAN management daemon governs policy, resource admission and durable outcomes through its private Unix-socket interface. The operator API is separate; the master coordinates but does not become the operator. The future SysOp and Security Officer daemons would be separate, permissioned services, not current Alpha functionality.