Hyperlocal Market Reports Plugin- Cowork Agent

What This Is

Sellers hire the knowledge broker. The National Association of REALTORS® Profile of Home Buyers and Sellers says it every year — the agent who can speak fluently about pricing, timing, and inventory in their area wins the listing. So why does your market update go out as “Nashville market update” to someone who lives in Brentwood?

This plugin fixes that. It turns your contacts and your MLS data into a branded market report email for every neighborhood in your database — one per ZIP, city, county, subdivision, or neighborhood — each one built on real numbers from that exact area. Median prices. Days on market. List-to-sale ratios. Inventory. And every email gives both sides of the story: what a homeowner’s house could fetch, and what a buyer in that area is up against.

Imagine 250 contacts spread across five ZIP codes. Today that’s either one generic email too broad to land, or five hand-built reports and a spreadsheet to figure out who gets which. This builds all five — plus a tagged recipient list so the right email reaches the right people — ready to import into Mailchimp, Constant Contact, ActiveCampaign, or SendGrid.

If a contact lives in Nashville but is searching in Franklin, they get the Franklin email too. And if your CRM doesn’t track where people are searching? It runs from home addresses and works just as well.

It writes files. It never sends. You review every email, import it into your email platform, and schedule it yourself. Your contacts and MLS data stay in a folder on your computer.

What’s different from the Cowork agent version

If you set up the earlier version of this agent, the engine is the same — the setup is what’s gone:

• No skills to install by hand. One plugin file, and every piece comes with it.
• No CLAUDE.md to place, and no config file to download and move. Setup writes everything straight into your campaign folder.
• No kickoff prompt to save. You just say what you want.
• Setup checks your folder first. It sees whether your contacts and MLS data are already there, and if your MLS export isn’t, it tells you exactly which areas to pull.
• It catches stale data. Run it with last month’s MLS export still in the folder and it tells you before it builds anything.

How to Use It

Setup takes about 10 to 15 minutes, once per campaign. Have these ready:

• Claude Desktop with Cowork. Campaigns run in Cowork, on your computer.
• A contacts export from your CRM, as a CSV. Follow Up Boss, Lofty, Sierra Interactive, or anything that exports a CSV works.
• An MLS export — active, pending, and closed listings from the last 90 to 180 days. Don’t know which areas to pull yet? Skip it. Setup tells you.
• An email platform that accepts pasted HTML emails. Mailchimp, Constant Contact, ActiveCampaign, and SendGrid all do. HubSpot’s drag-and-drop editor doesn’t — you’d need a coded email template set up first.

Step 1: Download the plugin

Step 2: Add it to Claude

Open Claude’s settings, find the plugins section, and add the file you just downloaded. That’s the whole install — no skills to toggle on, nothing to configure.

Step 3: Make a folder for this campaign

Every campaign gets its own folder. Create one on your Desktop or in Documents and name it for what it is — brentwood-sellers, williamson-luxury. Then drop in what you have:

• Your contacts CSV — save it as contacts.csv right in the folder, not in a subfolder.
• Your MLS export — inside a folder called mls. Don’t have it yet? Skip it.

brentwood-sellers/
├── contacts.csv
└── mls/
    └── your-mls-export.xlsx

Step 4: Set up your campaign

Open Claude Desktop, start a new Cowork task, choose your campaign folder, and say:

Set up my market reports

Setup looks in your folder first, so it isn’t asking about files that are already sitting there. Then it walks you through your campaign one short group of questions at a time — you can voice-dictate or just dump context:

1. Your campaign’s scope — its name, how to split your market (ZIP, city, county, subdivision, or neighborhood), any filters like price range or property type, and whether it leans seller, buyer, or balanced.
2. Your contacts — it reads your CSV’s actual columns and proposes how to map them, including where your CRM keeps buyer search areas. You confirm or correct.
3. The MLS data to pull — if your export isn’t in yet, it names the exact areas in your contacts, so you can pull them while you finish setup instead of exporting blind.
4. Your sign-off — name, title, brokerage, phone, reply email, and license number if your state requires it.
5. Your brand — colors, fonts, style, and your brokerage’s disclaimer if it has one.
6. Your design choices — corners, buttons, spacing, header, and dividers.
7. Your email platform — which one you use, how you organize recipients, and the tag it uses for a recipient’s first name.

Confirm the first-name tag. Every email opens with the recipient’s first name, and every platform writes that placeholder differently. Setup suggests the usual one for your platform — check it against your platform’s personalization menu. Get it wrong and the email literally says “Hi *|FNAME|*,” in your contact’s inbox.

Step 5: Build your campaign

If your contacts and MLS data are both in when setup finishes, it offers to build your first campaign right then. Otherwise, once your MLS export is in the mls folder, say:

Run my market report campaign

It shows you how it read your contacts, how many landed in each area after your filters, and whether your MLS data covers every one of those areas. It asks before it builds. Then it writes everything to a dated folder:

campaigns/2026-09-13-brentwood-sellers/
├── README.md         how to send it, for your platform
├── recipients.csv    every contact, tagged by area
├── segments/         one email per area
└── insights/         the numbers behind every email

Step 6: Send it

Open the README.md in that campaign folder. It walks you through your platform step by step:

1. Import recipients.csv one area at a time, tagging each import with that area’s tag.
2. For each area, open its email from segments, copy everything, and paste it into a new campaign using your platform’s custom HTML or code option.
3. Send each campaign only to its matching tag.
4. Send yourself a test first — and make sure the greeting shows a real first name.

Every run after that

Run it monthly, quarterly, or whenever your market moves. Export fresh contacts and a fresh MLS report, swap them into the folder, and say “Run my market report campaign.” Every run lands in its own dated folder, so you keep a clean history of every send.

Need to change your branding, filters, or sign-off? Open the folder in Cowork and just ask. Setup updates only what you name — no starting over.

One folder, one campaign

Each folder runs one focused campaign: one way of splitting your market, one set of filters. Want Brentwood sellers and Williamson County luxury buyers? That’s two folders, each set up once. Keeping them separate keeps every campaign tight — no 90-email runs you’d never send.

Tips

• Bring fresh data every run. The whole point is hyperlocal relevance. If your contacts or MLS files are more than 30 days old — or were already used by your last campaign — it flags them before it starts.
• Pull clean MLS data. Garbage in, garbage out. Missing prices or mislabeled ZIPs show up in the emails. A defined window of the last 90 to 180 days works best.
• Review every email before you schedule. On a small sample, AI can read a pattern into the numbers that isn’t really there. The emails soften their language on thin data, but you know your market.
• The compliance is built in. No agent names or rankings, no “great schools” or “safe neighborhood,” no invented numbers — every statistic comes from your MLS export, calculated in code. That’s not a limitation. It’s what separates professional marketing from liability.
• Keep your computer awake during a run. Cowork runs locally, so a sleeping laptop pauses the task. Turn on “Keep computer awake” in Claude Desktop’s settings for uninterrupted runs.

From the Session: This agent was introduced in Hyperlocal Market Report Campaigns.