Start here: Catalog Scout 1.0.3

What you receive:

a working new-arrivals checker: in your browser with no install, or as a Node batch tool;
line-sheet drafting;
drop-folder intake;
a client report;
a distinct brand;
researched positioning, editable economics, a first-customer plan, an AI operator prompt and a launch checklist.

This is a prototype plus a launch kit, not an operating business. Access and reuse are free and nonexclusive under LICENSE.txt.

Fastest path: no terminal (5 minutes)

Unzip. Everything is inside the catalog-scout-1.0.3 folder.
Double-click OPEN-ME-Catalog-Scout.html. It works offline, and nothing is uploaded.
Click Try the sample batch to check a fictional batch of eight décor products against its line sheet. You'll see the summary, "Fix this week", the product × fact grid and 10 facts drafted with row citations.
For real data: in Shopify admin, go to Products → Export → CSV for Excel, Numbers or other spreadsheet programs. Choose that file under Choose a product CSV, add the vendor's line sheet (a CSV with a SKU or Style # column), pick the product family and click Check this batch. Loading your own file clears the sample's store name.
Enter the client's store name (and, optionally, your business name), then click Download client report (HTML). Download Drafts CSV for the facts to paste after you verify them.
Keep the Structured report (JSON). Next batch, choose it under Compare with a previous report. Findings are matched by product handle and SKU, so added rows don't show up as changes.

Only check a store's catalog with the owner's permission.

Tested on Node 20.19.2 and 22.23.3. No npm packages, account, API key or purchase is needed.

In source/, run npm test. It runs 30 tests, including the mock Shopify pull.
npm run sample checks the sample with its line sheet and writes the report, CSV, JSON and drafts CSV into operations/. Run it again and it returns reused.
One client, by hand: node cli.mjs ~/Downloads/products_export.csv ../reports/fern-fable decor --client "Fern & Fable Home" --linesheet ~/Downloads/vendor-line-sheet.csv
Drop-folder intake (automated):
- Create clients/<name>/inbox/ and put exports and line sheets there. A line sheet needs "line sheet" or "linesheet" in its file name.
- Then run node intake.mjs clients/<name> --client "Store" --profile decor.
- It skips duplicates, moves handled files to processed/, and exits 2 with a "stale" status when no new export arrives within 8 days.
Scripted dry run: npm run dry-run rewrites operations/dry-run/LOG.md by running four simulated weeks.
Optional Shopify pull (tested against a local mock only): see source/pull-shopify.mjs. It needs an app the merchant creates in their own Dev Dashboard.
Before editing any detection rule: run npm run gate, and again after the edit (see AI-OPERATOR.md). npm run evaluate prints the accuracy tables in ACCURACY.md.

To host the checker yourself, serve source/ over http(s) and change the two BotMyBiz return links.

Top level: OPEN-ME-Catalog-Scout.html, BUSINESS-PLAN.md, RESEARCH.md, ECONOMICS.csv and ECONOMICS.md, FIRST-TEN-CUSTOMERS.md, AI-OPERATOR.md, LAUNCH-CHECKLIST.md, BRAND.md, ACCURACY.md, CHANGELOG.md and LICENSE.txt.
brand/: logo-light.svg, logo-dark.svg and cover.jpg.
source/: the demo, CLI, pipeline, intake, Shopify pull, line-sheet drafting, report, evaluator, gate, economics and tests, plus sample.csv, sample-linesheet.csv and the labelled fixtures.
operations/: the schedule example, a state template and the dry-run log.
package-manifest.json: a SHA-256 for every file.

Stated, not true: the checks tell you whether a fact is stated, not whether it is true.
Drafts: they are copied from the vendor file and are only as current as that file. Verify each one.
Not checked: image links are not opened, and currency is not verified.
Never reimport the report as a product catalog.
Not included: a live store connection for many clients, delivery, billing and a running scheduler. You set those up yourself.