The release gate
A script in your repo that refuses to release unless two different reviewers passed the exact commit that is about to deploy. Seven refusals, listed on the playbook page.
Harness setup · one fixed price · ten business days
The harness, installed in your accounts and handed to your team.
The release gate, the two-reviewer pipeline, the shared memory and rules, and the per-person budgets, set up remotely in your own repo on the AI plans you already pay for. Two real changes go through the whole path with your people pressing go. Then you own it.
Not sure yet? Start with the audit or Prefer to talk first? 20 minutes, free.
What gets installed
Four parts, in your repo, on your plans.
The two-reviewer pipeline
Two reviewer models from different families, called on your own AI plans, each returning a verdict file with the tree hash. The builder never reviews its own build.
Shared memory and rules
Rules, lessons and org facts as plain files in your repo. Every agent reads them before it works. A lesson written once applies to every agent from then on.
Per-person budgets
A daily spend limit for each person, checked before each model call. When it is reached, the call does not go out and the person is told why.
How the setup works
Ten business days, day by day.
Remote. About four hours of your team's time in total, most of it on days 1, 6, 8 and 9.
- Day 1
Kickoff and access
One remote call. Read-only invites to the repo and one sandbox. Which AI plan each person holds. The first rules file is written together.
- Day 2
The files
Rules, lessons and org-facts files in your repo. The brief template. The two reviewer skill files and the verdict template.
- Day 3
The gate
The release gate script installed and run against a test commit. Each of the seven refusals shown to fire, including an edit made after review.
- Day 4
The reviewers
Two reviewer models wired to your plans. First review run on a past change your team already shipped, so the findings can be checked against what really happened.
- Day 5
Budgets and checkpoint
Each person's daily limit set and the check placed before every model call. A planted overspend is stopped. Short checkpoint with the lead.
- Day 6
First real change
A brief, a build in a branch, two reviews, the gate, a check-only validation in a sandbox. Your named person presses go.
- Day 7
Post-deploy check and lessons
The post-deploy check for that change is written and run. The first lesson files come from what the two reviews found.
- Day 8
Second change, your hands
Your team runs the whole path with us watching. The brief template and the rules are fixed from what went wrong.
- Day 9
Training
Ninety minutes, recorded for you: how to write a brief, how to read a verdict, what the gate refuses and why, how to add a lesson, how to change a limit.
- Day 10
Handoff
The handoff document. Our access revoked. The written deletion confirmation scheduled within seven days. You owe nothing after this.
Before we hand off
Six things that have to be true.
Each one is shown, not described. The handoff document records when and on which change.
- The gate refuses a commit edited after review, and releases the same commit when the hashes agree.
- Two reviews from two different model families landed on a real change, each with the tree hash.
- A planted overspend was stopped by the per-person limit before the call went out.
- Your named person pressed go on a check-only validated job, and the post-deploy check passed.
- A lesson written by one person changed the next run of every agent.
- Your team ran the whole path once without us touching the keyboard.
What we need from you
A repo, one sandbox, your own AI plans, and one person who says go.
- A repo for your Salesforce metadata (GitHub, GitLab, Bitbucket or Azure DevOps).
- One sandbox for the two real changes. Never production.
- An AI coding plan for each person who will build or review, in your company's name.
- One named person who presses go, and about four hours of the team's time across the ten days.
How an engagement works
We set it up. We hand it back. That is the whole model.
Look
The audit: a read-only look at your repo and one sandbox, and a written playbook within 48 hours. If you would rather talk first, twenty minutes on a call, free.
Setup
Ten business days, remote, at one fixed price. The gate, the two-reviewer pipeline, the shared memory and rules, and the per-person budgets, installed in your accounts on your AI plans.
Handoff
Your team is trained and gets the keys. Nothing runs through StratoForce afterwards. You own every file, prompt, skill and memory, and we confirm our copy is deleted, in writing.
Advice, when you want it
One session at a time, 45 minutes, $199, booked when you need it. No retainer and no minimum. Most teams do not need it.
No retainer. No subscription. No usage fees. After handoff you owe us nothing unless you book another session.How we charge
Questions
Straight answers.
Is this the same as the Release Pack?
The files are the same. The setup installs them in your repo, wires the reviewers and budgets to your plans, runs two real changes through the path with your team, and trains them. The pack is free if you would rather do that yourself.
Do we have to do the audit first?
No. The audit is the cheaper way to find out which of the five rules you are missing. Teams that already know can start with the setup.
Which AI models does it use?
The ones on the plans you already hold. We need two reviewer models from different families; we work with whatever your company has approved.
What happens if a model is down during the setup?
The pipeline fails over to another model in the same role, and the builder still never reviews its own build. We show you the failover on day 4.
Can it touch production?
No. Nothing automated reaches production during the setup or after it. A person presses go, and only on a check-only validated job.
What if we need help later?
Book an advice session: 45 minutes, $199, paid once. If what you want is bigger, it is another fixed-price setup with its own written price. There is no retainer to be on.
Do you host anything for us?
No. Everything runs in your own accounts, on your own AI plans. At handoff you own every file, prompt, skill and memory, and we keep no copy. We confirm that in writing.
Start with the audit, or ask about the setup.
The audit is the cheaper first step and tells you which parts you are missing. If you already know, ask about the setup and we reply with the start date and the written price.