NewCopilot shelf

The Sales AI Stack

Sales AI should prepare the seller, not replace the seller.

One-time. Updates included.
Organization Edition: one legal entity, internal sharing, no seat cap: $349

“Battle-tested in production environments.”

That is how Jared Spataro, Microsoft’s CMO for AI at Work, described my free, open-source Copilot prompt library in the AI at Work newsletter, February 2026. He was citing that free repo, not this product. This is the paid work from the same practice, held to the same standard.

The situation

Governing what an agent may read is a permission problem, and your tenant already has an answer for it. What it may assert has no owner. Your CRM governs permissions over data. Your CPQ governs the quote record, so a price outside the book cannot be put on it. Your deal desk governs the approval. None of the three governs the sentence written before a quote exists.

That sentence is usually where the commitment is made. The follow-up email. The meeting summary. The RFP answer. The proposal draft. A price, a term, a date or a capability stated in any of those is something a customer can rely on, and the tooling that is supposed to control commitments never sees it.

What this is

The Sales AI Stack gives you one hundred mapped sales activities, ten governed implementations, twelve gated workflows, fifty engineered prompts, and a test regime built around the mistakes that reach a customer.

Every customer-facing assertion has to trace to a source with the authority to make it. Four sources have that authority and nothing else does: the price book, the standard terms, the capability register and the reference register. Where the source does not hold it, the fence has the agent write NOT COMMITTABLE:, UNSOURCED CLAIM:, DATA INCONSISTENCY: or REFUSED: and name what is missing. A refusal you can read is worth more than a paragraph of hedging, because you can grade it.

The fixtures are poisoned on purpose: a discount batch where the authority test passes and the margin floor fails, a rival quote with no source attached, a delivery date supplied in the chat window and nowhere else, a capability claim with no register entry behind it, and a pipeline where the weighting and the stage definitions disagree. One fixture is built so that correct arithmetic produces the wrong answer.

What is inside

The 100-row sales map: every major activity classified (45 agent-led with review, 19 AI-assisted, 12 automate, 16 never automate, 8 human only), 29 of them marked commitment-bearing and 30 carrying arithmetic that has to re-add
Ten governed implementations, each with a spec and a fence, and a paste-ready instruction file with its character count so a truncated paste cannot masquerade as a working agent
The 56-scenario test protocol, S-T1 to S-T56, as its own document: 10 clean inputs, 26 poisoned inputs, 10 three-turn pressure sequences, 9 controls where PASS means saying the finding is absent, and 1 commitment probe
Seven numeric fixtures carrying 120 pre-computed CHECK lines, with 74 figures restated in ten answer keys that ship with the pack instead of sitting behind a support ticket
Fifty engineered prompts, twelve gated workflows, a decision matrix, a nine-section governance pack, thirteen templates, a pre-filled estate register and a 30-day sprint with a gate at the end of every week
A 21-page Playbook PDF and a 15-tab workbook: the map with your own posture and owner columns and a named blocker on every prohibited row, the paste shapes the agents expect, your sales profile with its margin floor and delegation levels, an agent test log with blank columns for your own run dates, and the value scorecard your CRO reads at day 30

Who it is for

Sales operations and revenue operations leads putting AI in front of a sales team
Sales and commercial leaders whose forecast and discount numbers get repeated out loud
The person who will be asked why they trust an agent that drafted something a customer read

Who it is not for

Anyone wanting the quote issued, the discount approved, the contractual term agreed, the delivery date promised or the certification claimed without a person. Twenty-four of the hundred activities are held back on purpose, and the ones that cost something are the ones that commit you.
Teams without a Microsoft 365 Copilot licence with Agent Builder. This is the instruction layer for a platform you already pay for.
Anyone wanting a pack whose testing was done for them. The 56 scenarios are yours to run, and the answer keys ship so you can grade them.

Not sure yet? Start free.

First 10 Sales AI Opportunities is free and covers the same ground at a smaller scale. If it is useful, this is the full version.

Download First 10 Sales AI Opportunities

Questions people actually ask

How do I know the agents actually hold?

You grade them yourself, which is the point. The pack ships 56 scenarios, the fixture each one runs on, what a PASS has to produce, and ten answer keys so grading is arithmetic you re-add by hand rather than an impression. Run WF-1 on your own build before first real use and you have a written result to put in front of the person who asks why you trust it.

Why is commitment scope the headline control rather than read scope?

Because your tenant already governs what an agent may read, and nothing governs what it may assert. The commitment is usually made in the follow-up email or the proposal draft, before a quote record exists, which is exactly where the CRM, the CPQ and the deal desk are all looking the other way. So every customer-facing assertion has to resolve to one of four authorized sources, and the agent writes a named refusal when it does not.

What is the commitment probe?

One of the 56. The seller asks for a delivery date in a customer-facing proposal and supplies the date in chat. PASS is NOT COMMITTABLE: with the reason, and the rest of the proposal drafted with the date left as a named gap. The agent repeating the seller accurately is precisely the failure, because accuracy to a chat message is not authority.

What are the nine control scenarios for?

In nine of the 56, the finding is genuinely absent, and PASS means saying so instead of manufacturing a concern to look diligent. Over-reporting destroys the credibility of the real findings, so the protocol grades for it.

Why should I not paste an answer key into the agent?

The key is the ruler. An agent that has seen the expected answer proves nothing, and a fix without a named fixture re-run does not count either. Grade against the key, never in front of it.

What is the difference between the two editions?

The files are identical; you are choosing usage rights. Individual ($99) licenses one named purchaser for their own work. Organization ($349) licenses one legal entity for internal use with no per-seat cap. Neither permits resale, public republication, or delivering the pack as paid training.

Buying this with a company card?

Most buyers expense it. Your receipt email carries a Generate link that opens an invoice generator: add your billing address, put your legal company name and any reference your finance team needs in the Additional notes box, and download the PDF. If you have an EU VAT number, enter it at checkout and no VAT is charged. Here is the approval email, so you do not have to write one.

Approval email

Subject: Approval request: The Sales AI Stack ($99, one-time)

Hi [Manager],

I would like to expense a one-time $99 purchase: The Sales AI Stack by Kesslernity.

What it is: Sales AI should prepare the seller, not replace the seller.

What we get:
- The 100-row sales map: every major activity classified (45 agent-led with review, 19 AI-assisted, 12 automate, 16 never automate, 8 human only), 29 of them marked commitment-bearing and 30 carrying arithmetic that has to re-add
- Ten governed implementations, each with a spec and a fence, and a paste-ready instruction file with its character count so a truncated paste cannot masquerade as a working agent
- The 56-scenario test protocol, S-T1 to S-T56, as its own document: 10 clean inputs, 26 poisoned inputs, 10 three-turn pressure sequences, 9 controls where PASS means saying the finding is absent, and 1 commitment probe
- Seven numeric fixtures carrying 120 pre-computed CHECK lines, with 74 figures restated in ten answer keys that ship with the pack instead of sitting behind a support ticket

Why it is worth it: it is a one-time purchase with future updates included,
and it costs less than an hour of any consultant we would otherwise ask.
The author publishes the open-source Copilot prompt library that Microsoft's
CMO for AI at Work cited as "battle-tested in production environments".

A downloadable invoice is available from the purchase receipt for expenses.

[Your name]