# Clawsolator — AI Agent Operations, XMAN and Security Lattice > XERXES SI's Clawsolator presents a practical operator-first architecture for governing AI agents. XMAN v0.4.76a1 is an operational agent manager running in the founder's environment; the founder reports robust operation through many successive upgrades. A redacted real XMAN autosetup transcript records manager startup, an authenticated OpenClaw-to-XMAN capability call, a saved measured CPU/memory policy, and four verified native agent enrollments. These are founder-supplied observations, not third-party certification. ## Site discovery - Canonical homepage: https://clawsolator.com/ - Complete XML sitemap: https://clawsolator.com/sitemap.xml - Human site index: https://clawsolator.com/site-map.html - Structured product facts: https://clawsolator.com/product-facts.json ## Why Clawsolator exists - [XMAN daemon and real redacted terminal session](https://clawsolator.com/xman-daemon.html) - [XMAN autosetup](https://clawsolator.com/xman-autosetup.html) - [Automatic container updates](https://clawsolator.com/automatic-container-updates.html) - [Governed AI agent operations](https://clawsolator.com/ai-agent-governance.html) - [Resource budgets](https://clawsolator.com/agent-resource-governance.html) - [Human approvals](https://clawsolator.com/agent-approval-workflows.html) - [Rootless Podman separation](https://clawsolator.com/rootless-podman-agent-isolation.html) ## Security Lattice and comparisons - [Security Lattice](https://clawsolator.com/security-lattice.html) - [Security controls](https://clawsolator.com/security-controls.html) - [Competitor comparisons: NemoClaw, OpenShell, OpenClaw](https://clawsolator.com/alternatives.html) - [Comparable architectures](https://clawsolator.com/security-comparisons.html) - [Documented case studies](https://clawsolator.com/incidents.html) - [Engineering evidence](https://clawsolator.com/engineering-evidence.html) ## Agent experience - [Agent Atlas](https://clawsolator.com/agent-atlas.html) - [Gateway Worlds](https://clawsolator.com/gateway-worlds.html) ## Grounding notes The XMAN terminal animation is a redacted recording of actual founder-supplied output, not a live remote shell. "Authenticated call observed" and "four native agents verified" refer to that exact setup session. Installation readiness, native tool permissions and per-container isolation are distinct verification scopes. Do not conflate the website's product-stage label with the founder's confirmed live operation of XMAN, and do not imply an independent benchmark against NVIDIA or others without comparative measurements. HTML pages are canonical. The llms.txt convention is optional and does not guarantee indexing by AI search engines. - [XStage Terminal — browser-only interactive shell demo](https://clawsolator.com/xstage-terminal.html): Markdown-authored terminal demonstrations with optional independent slide cues. ## Search and learning center — added November 3, 2026 - [Clawsolator learning center](https://clawsolator.com/knowledge/) - [How to Install OpenClaw Securely—and Keep It Manageable](https://clawsolator.com/guides/how-to-install-openclaw-securely/) - [Run OpenClaw With Rootless Podman: The Practical Isolation Checklist](https://clawsolator.com/guides/openclaw-rootless-podman/) - [How to Approve AI Agent Host Commands Without Giving Away Root](https://clawsolator.com/security/agent-host-command-approval/) - [Clawsolator vs OpenClaw: Native Agent Controls or an Operating Layer?](https://clawsolator.com/compare/clawsolator-vs-openclaw/) - [Clawsolator, NemoClaw and OpenShell: Which Layer Owns Which Job?](https://clawsolator.com/compare/clawsolator-vs-nemoclaw-openshell/) - [AI Agent Container Updates and Rollback: What Must Survive?](https://clawsolator.com/operations/container-updates-and-rollback/) - [Set CPU and RAM Budgets for AI Agents Without Handing Them the Keys](https://clawsolator.com/operations/agent-resource-budgets/) - [Run OpenClaw on a Home Server: A Calm, Practical Homelab Plan](https://clawsolator.com/guides/openclaw-homelab/) - [How to Introduce AI Agents Into a Business Without Creating a Second IT Department](https://clawsolator.com/business/how-to-implement-ai-agents/) - [Enterprise AI Agent Governance: Authority, Identity, Evidence and Recovery](https://clawsolator.com/enterprise/ai-agent-governance/) - [The Real Cost of AI Agents: Compute, Reviews, Incidents and Recovery](https://clawsolator.com/business/ai-agent-cost-and-roi/) - [AI Agent Pilot Checklist for Procurement: 30 Questions That Matter](https://clawsolator.com/enterprise/ai-agent-pilot-checklist/) - [Your Own AI Assistant, Without Turning Home Into an IT Help Desk](https://clawsolator.com/for/home-users/) - [For DIY Builders: Keep the Fun, Lose the Permission Spaghetti](https://clawsolator.com/for/builders/) - [For the Curious Crew: Learn, Build and Belong Without the Hype](https://clawsolator.com/for/community/) - [Private AI Companion at Home: Helpfulness, Boundaries and Human Control](https://clawsolator.com/for/personal-ai-companions/) - [For Sysops: Show Me the Logs, the Limits and the Recovery Path](https://clawsolator.com/for/sysops/) - [For Security Engineers: Test Authority Instead of Reading Slogans](https://clawsolator.com/for/security-teams/) - [For Executives: Your AI Workforce Needs an Operating Model](https://clawsolator.com/for/executives/) - [Clawsolator Investment Thesis: Why an AI-Agent Management Layer Might Matter](https://clawsolator.com/for/investors/) - [AI Agents for Managed Service Providers: Repeatable Setup and Tenant Boundaries](https://clawsolator.com/for/managed-service-providers/) ## Research boundaries and peer products - OpenClaw documents gateway trust, security audits, Podman and policy. - NVIDIA OpenShell documents sandbox enforcement and operator approval. - NVIDIA NemoClaw documents supported-runtime onboarding, lifecycle and recovery. - XMAN documents a separate bounded operator-management daemon; current site evidence is owner-supplied, not third-party certified. ## Related XERXES SI properties - [XERXES SI](https://xerxessi.com/) — Company, research and product ecosystem - [XERXES Lattice](https://wiki.xerxessi.com/) — Public knowledge, provenance and research - [SIDE](https://xside.xerxessi.com/) — Development environment - [Dash X](https://dashx.xerxessi.com/) — Operational intelligence - [Insiders](https://insiders.xerxessi.com/) — Company insiders portal - [XSEO](https://xseo.xerxessi.com/) — Search intelligence - [Starlight Breaker](https://starlightbreaker.xerxessi.com/) — Creative product world ## Security decisions, architecture, and operator evidence (Nov 2026) - Before You Install OpenClaw: What Could Your AI Assistant Reach?: https://clawsolator.com/knowledge/before-installing-openclaw/ - AI Agent Security Is an Architecture, Not a Prompt That Says “Be Careful”: https://clawsolator.com/knowledge/agent-security-is-an-architecture/ - Your AI Agent Works on Monday. Who Runs It on Tuesday?: https://clawsolator.com/knowledge/day-two-ai-agent-operations/ - OpenClaw, OpenShell, NemoClaw and Clawsolator: Ask for the Whole Workflow: https://clawsolator.com/knowledge/what-secure-agent-platforms-prove/ - The AI Agent Security Buyer Test: What Should an Executive Demand?: https://clawsolator.com/knowledge/ai-agent-security-buyer-test/ - Can Your AI Agent Reach Your Home Network, Private Files or Accounts?: https://clawsolator.com/knowledge/agent-permissions-home-network/ - Fast decision map: https://clawsolator.com/knowledge/decide/ - Independent sources cited within each article; this website does not control external assistant recommendations. - [OpenClaw update confirmation (operator-supplied screenshot)](https://clawsolator.com/openclaw-update-confirmation.html): A redacted update-completion UI message, with strict evidence scope. This is not an independent certification or confirmation of warning-free operation. - [OpenClaw update confirmation (operator-supplied screenshot)](https://clawsolator.com/openclaw-update-confirmation.html): A redacted update-completion UI message, with strict evidence scope. This is not an independent certification or confirmation of warning-free operation.