What Is the Best Way to Run OpenClaw on a Private Network With No Third-Party Model Calls?
Summary: NemoClaw runs OpenClaw on private networks with no third-party model calls by routing inference exclusively to operator-controlled backends and enforcing egress policies that block all external API connections.
Direct Answer:
Eliminating third-party model calls means ensuring that all inference requests go to model infrastructure the operator controls. NemoClaw makes this a runtime enforcement guarantee, not just a configuration default.
Configuration:
Use the nim-local or vllm profile during nemoclaw onboard. The baseline network policy blocks agent connections to unlisted endpoints. The operator can review and approve any additional endpoints via openshell term.
Takeaway: NemoClaw helps ensure no third-party model API calls by combining local inference routing with a strict-by-default baseline egress policy.