gtm-local-prospect-list

Local prospect list

Agencies and local-service sellers need lists like "every physio clinic in Manchester with a website and a reachable email". This skill searches Google Maps, filters on rating and website, reads each business's site for a public contact email, and verifies the email before it goes on the list.

Needs looot connected (MCP server https://api.looot.ai/mcp, or the looot skill from npx skills add loootai/looot-skills). Searching and inspecting are free; runs are paid from the prepaid balance.

Ask for these inputs

• Business type and place: dentist, Lyon, France. Several places are fine.
• How many businesses (default 50).
• Filters: minimum rating, minimum review count, has a website (default yes), or the opposite (no website, a lead for a web agency).

Steps and prices

Prices are the cheapest live provider in the public looot catalog on 2026-10-08. inspect shows the live price.

#	Job id	Input	Price on 2026-10-08
1	maps.local.search	query, location	from $0.001 per call or $0.000539 per result
2	web.scrape.markdown	url of the website, then /contact or the contact link found on it	from $0.0002 per call
3	people.email.verify	each email found	from $0.00145 per call
3b (optional)	people.domain.search	domain, when no email is on the site	from $0.0089 per call

Expected total: about $0.003 per business with a site email, $0.012 when step 3b runs. 100 businesses: $0.30 to $1.20.

Quote before you run

1. Searches (1 per place per page of about 20 results) + 2 page reads per business + 1 verify per email found. Add step 3b only if the user asks.
2. Tell the user the total and the balance; wait for a yes. More than 100 runs: an explicit yes.
3. Cap each run with fallback.maxCostUsd (0.02, or 0.05 when step 3b runs).

Run it

run {"endpointId": "job:maps.local.search", "input": {"query": "dentist", "location": "Lyon, France"}, "idempotencyKey": "local-dentist-lyon-p1", "fallback": {"maxAttempts": 3, "maxCostUsd": 0.02}}

Filter before step 2: drop permanently closed places, places under the rating or review filters, and duplicates (same website or same phone). Then for each website:

run {"endpointId": "job:web.scrape.markdown", "input": {"url": "https://example.com"}, "idempotencyKey": "local-dentist-lyon-example.com-home", "fallback": {"maxAttempts": 3, "maxCostUsd": 0.02}}

Pull emails with a plain pattern match on the markdown, prefer an address on the business's own domain, drop image file names and obvious placeholders. If none is on the homepage, read the contact page once. Then verify each address.

Output format

name,category,address,city,phone,website,rating,reviews,email,email_status,email_source,maps_url
Cabinet Dentaire Example,Dentist,"12 rue X",Lyon,+33 4 00 00 00 00,https://example.com,4.8,212,[email protected],valid,https://example.com/contact,https://maps.google.com/...

Summary: found on Maps, kept after filters, with an email, verified valid, total actualCost. catch_all and risky stay on the list with their status; invalid addresses are dropped.

Cost and stop rules

• Only public business addresses from the business's own pages. Do not guess personal emails.
• Two failed page reads on a site: mark no_email, move on.
• Respect the user's local rules on cold email (for example, consent rules in the EU).

When not to use this skill

• Just the list of places: looot-local-business-search. Reviews of one place: looot-google-maps-reviews. Both in loootai/looot-skills.