Agency Preview
Public Preview · Agency Edition
For advisors, consultants, and boutique firms delivering AI governance to clients.
Real pages, real instruments, and real operator guidance from the Agency Edition — the professional deployment package licensed to advisors who bring the Decision Protocol to their engagements.
- Excerpts from the two core governance frameworks
- Sample pages from the Operator's Manual
- Preview of three decision file tiers
- One worked case record from the Execution Library
- Firm-branding permissions and license scope
Early Access $997 · Standard $1,497 from 1 May 2026. Identical contents and license scope. Only the acquisition price differs.
Your clients are asking for AI governance. The gap is operational, not conceptual.
Boards are asking what the AI policy is. Regulators are asking to see the decision record. Clients are asking their advisors the same question: "Help us show that we governed this." Generic policy language does not answer that question. What answers it is structured documentation produced at decision time — and a repeatable method your team can run again on the next engagement.
The Agency Edition gives you that method, in a form you can deploy under your own firm brand without building it from scratch.
- Each engagement re-invents the scoping conversation.
- Policy language looks generic; clients cannot hand it to the board.
- Triage is ad hoc; decisions are recorded inconsistently.
- No audit trail when the regulator arrives.
- Difficult to delegate the work inside the firm.
- Scoping memorandum runs the first meeting.
- Firm-branded deliverables go straight to the client.
- Triage is structured; three decision-file tiers fit the risk.
- Every decision produces a defensible record, dated and signed.
- Repeatable playbook. New engagements start faster.
The shift is from "we advise on AI governance" to "we deploy a method — and we can show the record."
Five layers, one deployable package.
Two frameworks that solve two different internal control problems.
The method separates two control problems that clients tend to conflate: how people inside the organization may use AI, and how external AI tools may be introduced to the organization. The two frameworks address these separately, but are designed to function as one deployable governance layer.
- Acceptable use — permitted, restricted, prohibited.
- Literacy readiness before tool access.
- Shadow AI disclosure — time-limited, no-retaliation.
- Use-case triage by risk tier.
- Rollout memo and acknowledgment.
- Control ownership & approval matrix.
- Classification guide and executed example.
- AI vendor & data control policy.
- Vendor review checklist — six assessment domains.
- AI tool onboarding protocol.
- Restricted-data triage tool.
- Prompt & data handling rules.
- Vendor escalation & approval matrix.
- Executed example (SaaS tool approval).
Each is deployable on its own. Together they form a coherent internal governance layer.
Inspect the actual writing standard — six pages from the Agency Edition.
These are real pages from the Agency Edition materials, rendered here with a preview watermark. The full package ships without watermark and in editable source format where relevant.
Internal AI use requires more than permission. It requires structure.
The AI Acceptable Use Policy is a short, operational document — not a legal disclaimer. It tells your client's employees, in plain language, what may be done with AI, what may not, and what requires approval. The three tiers below are excerpted from §3 of the template in the Agency Edition.
- Full policy source in .docx — editable for your firm's client base
- Matching editable HTML working file
- Companion Classification Guide explaining the use-tier logic
- Executed example showing a completed policy deployment
Classification matters before escalation does.
A governance framework is only useful if your client's team can determine, quickly and consistently, what level of review a given AI use case requires. The triage tool routes any internal AI use case through three short questions, and resolves it into one of four outcomes. Try it below with an advisor-oriented example.
The same routing logic ships as an editable HTML tool your client's team can run directly.
Every engagement begins with a scoping memorandum.
The Scoping Memorandum is the first template in the editable source pack. It is short by design — one to two pages, signed by both the advisor and the client sponsor — and it establishes what is in scope, what is out, who decides, and what the engagement is not. Fill the fields below to try the logic.
The template is in editable .docx — customize once for your firm, reuse on every engagement.
How the advisor actually runs the engagement.
The Operator's Manual is the piece of the package that does not exist in the executive edition. It is the 28-page guide for the professional running the engagement: how to scope it, who to put in the room, how to facilitate the triage, how to handle the three most common client objections, and how to hand the governance output back without leaving your firm exposed. One chapter excerpted below.
A triage session is not a discovery workshop. By the time the advisor arrives, the organization has already started using AI — often in ways the board has not formally approved. The session's purpose is to surface that activity, classify each use case, and produce a decision record the organization can defend.
Three practical points set the tone:
- Run the scoping memorandum through the client sponsor before the session, not during it. The session itself must start with items already on the table, not with a blank sheet.
- The triage owner in the room should be a control-side function (compliance, risk, legal, information security) — not the team that is already using the tool. Separation of concerns matters even at the facilitation level.
- Every case surfaced in the session produces a decision file. Express, Standard, or Shielded is decided in the room, with the time allocated to each file disclosed up front.
- Scoping Memorandum playbook (Ch. 2)
- Stakeholder mapping for governance engagements (Ch. 3)
- Objection handling: "we already have a policy" and two other variants (Ch. 5)
- Engagement handoff and transfer memorandum (Ch. 7)
A worked case the advisor can use to calibrate the method.
The Execution Library contains 20 fully worked case records — trigger event, stakeholders, classification logic, the decision with its rationale, trade-offs recorded, and the instruments used. It is the component that transforms a document pack into a repeatable method. Case #11 is shown below in abbreviated form.
Twenty cases across AI use, vendor, and integrated decisions. Each one is a calibration point.
How the method moves through a client engagement.
The package is engineered so that a professional can move from first client meeting to the first defensible governance record without external support. The typical arc is below. The instrument column identifies which pieces of the package carry the work at each stage.
Agency Edition · Early Access
- 56 files across two core frameworks, bundle components, operator materials, and editable templates.
- Single-user perpetual license · firm-branding permissions on client deliverables.
- Method updates through 31 December 2026.