PRODUCT.md
Messages stop describing features and start describing what changes for the person receiving them. Written for the assistant that has to sell it, not for a product page. The sections that matter are the ones product marketing usually omits: what it does not do, what it replaces, and what a sceptical buyer says first.
# Product — <Product name>
## The outcome
<Multi-entity payroll close in two days instead of nine.>
## The mechanism
<It reconciles inter-entity payroll postings automatically by matching on employee-period-entity keys, so the manual matching that grows with each new entity stops growing.>
One sentence on HOW, because a claim with no mechanism reads as marketing and gets ignored by exactly the people who could buy it.
## What it replaces
<Three spreadsheets, two part-time weeks of a payroll manager's month, and the reconciliation step the auditors always query.>
## What it does not do
<- Does not run payroll. It reconciles what your payroll systems produced.
- Does not integrate with SAP.
- Does not handle US payroll.
- Does not remove the need for a payroll manager. It removes about forty per cent of their close week.>
This section is not a weakness. An assistant that knows it writes messages that survive the first reply.
## Who it is bought by, who it is used by
- Signs: <Finance Director>
- Uses: <Payroll Manager>
- Can block: <Head of IT, on the data extract>
These are frequently different people with different objections. Address the user's pain to the user and the risk to the buyer.
## Proof
| Claim | Evidence | May we name the customer? |
|---|---|---|
| <Nine days to two> | <Named customer, written consent held> | <Yes> |
| <40% of close week returned> | <Two customers, measured over 3 months> | <No — numbers only> |
## Pricing posture
<Per entity per month, from <price>. We publish it. Do not quote a discount, ever — that is a conversation with a human.>
## Objections it attracts
<See OBJECTIONS.md.>
## What makes a bad fit
<Single-entity companies. Companies mid-way through an ERP migration — they will not add a system during one, and saying so early earns more credibility than pushing.>- Claude Project knowledge
- ChatGPT Project files
- Agent folders
- All sections. "What it does not do" is the one to spend time on.
Everything in <angle brackets> is example content, not instructions. Replace it. The file ships filled in rather than empty on purpose: an empty template gets abandoned, a filled one gets edited.
The reference behind this file
A worked example of the pricing posture question: what it looks like when the numbers are on the page with the date they were checked.
Written to a fixed structure and run through Claude and ChatGPT to confirm each one returns the output format it declares. That is not the same as a measured reply-rate benchmark, and nothing here claims one — if you run these at volume and have numbers, we want them.
Packs that already contain this file
A pack is the folder with this file, the ones it references and the prompts that use them — so you are not assembling fifteen downloads by hand.
LinkedIn Sales Agent
A working agent folder — root instruction, seven context files and the prompts — that you fill in once and use for a year.
11 files, 8 promptsLinkedIn Content Agent
Drafts in your voice, built from real material, with a self-critique attached before you read them.
6 files, 5 promptsLinkedIn AI Starter Pack
The whole toolkit as one folder, set up in twenty minutes, whatever your role is.
12 files, 12 promptsLinkedIn MCP Setup Kit
A folder you can wire an assistant to LinkedIn from, with the exposure of each route stated before the first run rather than after it.
4 files, 3 promptsOne file is a start. The folder is the point.
Files reference each other — CLAUDE.md declares the reading order, the prompts expect the files by name. The packs ship them together, already wired up.