The Operator's Manual — C2 Command Layer

The capstone above all six pillars — not a seventh pillar, the seat that conducts them. The managed C2 command layer is the done-for-you version.

Six capabilities are only a department if one seat conducts them. Compliance, DevSecOps, PTaaS, AI-SOC, incident response and the fractional CISO each produce signal. Without a single place to see and direct them you have six tools and no department. The operator seat is what turns capability into coordination.

The C2 console — one pane, the whole department. An intake queue where reports from every doorway land, triaged and ready for action behind a gate. Approval gates, where every consequential action — a containment, an offensive re-test, a patch merge — waits for a human click that is visible and logged. The living risk register, where findings from every arm become tracked, closing, audit-ready entries. And AI-channel observability: volume and cost per channel, prompt-injection and jailbreak watch, and escalation flow, so the autonomy is never a black box.

Four doorways, one intake. Wherever an employee already works, the security department is one click away — an intranet portal, a browser companion, a desktop tray app and an Outlook add-in, all feeding the same queue and the same AI gateway. Reporting a phish, requesting access or asking a question all land in the operator seat. Because the doorways are first-party, you can swap your SIEM or MDR and the department keeps its visibility.

Grounded in one knowledge base. An Obsidian-git vault plus IaC JSON holding assets, topology, personas, runbooks, findings and history in one schema, version-controlled for a full audit trail. Every arm reads and writes it — a pentest finding, a triaged alert and a compliance gap all land in the same vault and update the same register. It learns by accretion rather than retraining: institutional memory that grows with every engagement. And zero-trust segmentation means the AI layer can be severed in one click with production untouched — the containment guarantee.

The agent layer. Run it on an agent framework — OpenClaw, Hermes, Grok- or Claude-based — held to three invariants: one vault, one gate, full observability. The guide covers wiring it to the CISO Marketplace MCP and API data feeds, where an agent can assess posture, write findings back to your reports, source remediation across 170+ services, and draft a proposal into a Finance pipeline it is never allowed to price.

Agent wallets, and why spend becomes a control. The 2026 shift is that agents hold wallets and spend autonomously, which turns budget into a gated security control rather than a finance concern. The guide treats it that way.

On a box, if you want sovereignty. The same pattern runs on hardware you hold — Qubes laptops and Linux NUCs with hardware tokens — the approach Pillar 04's SOC in a Box takes for the SOC, applied to the operator seat.

Honest limits, stated plainly. A console, not an autopilot: it conducts and surfaces, the operator directs, and the gates exist precisely so a human stays in command. It conducts your tools rather than replacing them — it orchestrates what you already run. It is built to run solo, with a contract analyst slotting into the triage tier as volume grows and no re-architecting required.

Includes the operator console reference layout, a fillable approval-gate policy for deciding in advance what needs a human click, the knowledge-base schema with the sever procedure, an MCP connection starter, and a full operator readiness checklist.

Buy it with the pillars it conducts — it is priced as part of a programme rather than sold as an island.