ENGAGEMENT & PRICING — THE GATEWAY LADDER

Create a free workspace. Step up to the mission boundary.

The free plan is available for up to 15 users. The AI Chat Firewall extension is available now as a managed pilot. The Mission Authorization Gateway is briefing-led: assessment, test-range pilot, then production per enforcement point — not per seat.

FLAGSHIP · HIGH-ASSURANCE TIER
Mission Authorization Gateway
For mission-critical and safety-critical boundaries — cross-domain, DDIL, OT, and air-gap-capable architectures. Designed against NSA cross-domain standards. Engaged via technical briefing and a scoped pilot.
ASSESSMENT: FIXED PRICE · REMOTE PILOT: TIME-BOXED · CREDITED TO YEAR ONE PRODUCTION: PER ENFORCEMENT POINT — NOT PER SEAT
Request a 30-minute Boundary Review
Bring one AI workflow, one consequential action, or one data boundary.
See the full ladder
The engagement ladder: assessment, test-range pilot, production, edge/OT. Four ascending rungs — a fixed-price remote assessment, a time-boxed test-range pilot whose fee is credited in full toward year-one production, production priced per enforcement point (annual, not per seat), and a custom-quoted edge/OT rung for DDIL and air-gapped boundaries. Engagements are sized by enforcement point and mission, not seats. No prices are shown.
FIG. A — THE ENGAGEMENT LADDERCONTAINMENT.AI
THE GATEWAY LADDER — FOUR RUNGS, ONE PROCUREMENT STORY
01 · ASSESSMENT
Fixed price
REMOTE · TIME-BOXED
Architecture and policy-surface review of your agent and autonomy stack. Written findings, plus a run of our traceability corpus against a representative policy set.
Scope an assessment
02 · TEST-RANGE PILOT
Time-boxed
FIXED FEE · 100% CREDITED TO YEAR ONE
The Gateway deployed on your own unclassified or simulation network. Policy enforcement runs shadow-first, then enforcing. You receive an evaluation report with the full trace evidence.
Plan a pilot
PRODUCTION
03 · PRODUCTION
Per enforcement point
ANNUAL · NOT PER SEAT
An enforcement point is one deployed Gateway decision point governing one agent fleet or system boundary. Metered by enforced actions, not seats. Multi-point pricing is set out in your proposal.
Request a 30-minute Boundary Review
04 · EDGE / OT
Custom
SCOPED AND QUOTED AT BRIEFING
Disconnected, intermittent, and operational-technology boundaries — DDIL and air-gap-shaped deployments are scoped against your mission profile.
Start the conversation
Pricing for each rung is provided in capability briefings and proposals. The ladder is shaped so a 2-year, single-enforcement-point engagement — net of the pilot credit — fits one simplified-acquisition purchase order.
FREE WORKSPACE — AVAILABLE NOW · EXTENSION MANAGED PILOT AVAILABLE
$0 UP TO 15 USERS
Create the workspace now. Browser enforcement begins when the extension is installed through an approved distribution path.
✓ AI Chat Firewall for ChatGPT, Claude, Gemini, Copilot, and Grok
✓ Prebuilt guardrail library
✓ 30-day audit logs
✓ Email support
Create a free workspace Request a managed pilot
WHICH RUNG IS YOURS?
"We want to see the discipline work on our own team first." FREE → "We need a written, expert read on our agent and autonomy stack." ASSESSMENT → "We want the Gateway proven on our own test range before we commit." TEST-RANGE PILOT → "We're ready to govern a live boundary." PRODUCTION → "Our boundary is disconnected, intermittent, or OT: cross-domain, DDIL, or air-gap-capable." EDGE / OT BRIEFING →
HOW TO BUY — CONTRACTING IDENTITY
The identifiers a contracting officer needs to look us up and award. Current registration status is authoritative in SAM.gov — verify it there, not here.
LEGAL ENTITYContainment.ai LLC
SAM.GOV UEIVTC6UZQ27PH4
CAGE CODE16ZL7
PRIMARY NAICS518210 — Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services
PLACE OF PERFORMANCEGreat Falls, VA 22066, United States
AWARD SHAPESA full two-year, single-enforcement-point engagement — net of the pilot credit — is shaped to be awardable as one simplified-acquisition purchase order. Larger or multi-point scopes go through a capability briefing and a written proposal.
Current registration status is authoritative in SAM.gov. Certification and authorization status is stated in full on Trust. For contracting questions: PARTNERS@CONTAINMENT.AI.
COMMON QUESTIONS
Can we create a free workspace and grow into the Gateway?
Yes. Create the workspace now; add browser enforcement through an approved extension distribution path, then step onto the assessment, pilot, and production ladder as the boundary hardens.
What does the test-range pilot include?
The Gateway deployed on your own unclassified or simulation network, enforcing policy in a shadow-first, then enforcing sequence, with an evaluation report that carries the full trace evidence. The pilot fee is credited in full toward year-one production.
What is an enforcement point?
One deployed Gateway decision point governing one agent fleet or system boundary. Production is priced per enforcement point and metered by enforced actions — we do not price per seat.
Can we buy a rung on a purchase card?
No — assessments and pilots are services, and we do not structure buys to shave procurement thresholds. Instead, the ladder is shaped so a full two-year, single-enforcement-point engagement — net of the pilot credit — can be awarded as one simplified-acquisition purchase order. Exact figures are set out in your proposal.
Are you certified for government use?
The Gateway is designed against NSA cross-domain standards, and we run a certification roadmap covering SOC 2, ISO 27001, and FedRAMP/ATO pathways. Live milestone status is on the Trust page — the canonical record.
Does a model ever make the enforcement decision?
No. On every rung, the ruling is deterministic policy evaluation — same input, same ruling, replayable. Models never hold the gavel.