The SR 26-2 Gap Check

I want to set this up properly so Claude checks my model risk program against the new SR 26-2 rule on my real inventory, not a one-off chat. Walk me through it step by step, do not skip anything. Treat me like a model risk officer, CRO, or validation lead who has never built a Claude Project and does not write code. Define every term once.

This is NOT a Terminal or coding install. There is nothing to compile. It is a private Claude Project where the model inventory and governance docs live.

## The setup — a private Claude Project
Walk me through ONE step at a time, waiting for me to confirm each:

1. **What I need.** A Claude account on a Team or Enterprise plan so the data stays in my tenant. Pin **Opus 4.8**. Nothing to install.
2. **Make the workspace.** In Claude, create a new **Project** named for the institution and the review. A Project is a private workspace with its own knowledge other chats cannot see. My inventory and reports stay in my own tenant.
3. **Load the records.** Drop the model inventory, the validation reports, and the model risk governance policy into the Project knowledge. A spreadsheet inventory is fine.
4. **Run the vault.** Run prompt 01, pick the scope (full readiness, one model family, the vendor models, or a quarterly refresh) and where the data is. Then run 02 to 08: map to SR 26-2, find the shift-left gaps, re-grade the tiers, catch the vendor models, the self-check, the remediation plan, and the examiner-ready file.
5. **Gate it.** Treat prompt 06 (the self-check) as a hard stop. A **GATE: FAILED** means the two passes disagree on a gap; reconcile before continuing. This is a first-pass readiness review, not a filing. A named risk officer signs; legal and risk own the final program.

## Rules for walking me through this
- One step at a time. Define every term once: Project, knowledge, SR 11-7, SR 26-2, model inventory, validation, shift left, model tier, third-party and nth-party model, regulated-decision model, review gate.
- Do NOT tell me a step is "not possible." If I cannot find the Project button, it is in the left sidebar of claude.ai.
- Never paste supervised or confidential data into a public chat. Use a Project with my tenant controls.
- This maps the new rule to my inventory and finds the gaps. It does not validate a model or replace my validation function. Not legal or regulatory advice.

First message: ask me "What is the scope: full SR 26-2 readiness across the inventory, one model family, the vendor models, or a quarterly refresh? And do you have the model inventory and governance policy ready to upload?" Then start step 1.